GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
GIS training for companies
With more than 150,000 hours of training, GeoPoint, in addition to its e-learning offer, offers the possibility of carrying out GIS training actions tailored to its clients, aimed at companies or pre-organized groups.
We can adapt any content to your needs!
As a standard offer we have GIS courses in the following areas:
- Introduction to GIS
- Introduction to QGIS
- Management of Spatial Databases
- Editing, Data Processing and Spatial Vector Analysis
- Matrix Spatial Analysis – Raster
- 3D Spatial Analysis
- GeoMarketing
- Toponymy and Addresses
- WebGIS
GIS training in e-learning
GeoSDM (Geo Susceptibility Distribution Models) graduation
Description: professional graduation for Species Distribution Modeling (SDM) for ecological restoration; Surface motion monitoring (InSAR) for mineral exploration and mining sites monitoring; Geosimulation of LULC Changes to assess of the Landscape and habitats structure and dynamics; and Flood Detection based in EOD data feeds and Geographic Information Technologies (GIT) & Artificial Intelligence (AI). GEOSDM includes the following modules (you may ).
You can attend the courses individually or all at once sequentially.
Students are expected to have knowledge of GIS and at least basic knowledge of QGIS and Python.
- GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
- GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
- GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
- GeoSDM module 4– Smart Flood detection with GeoAI in Python.
GeoSDM module 1 – Species Distribution Models (SDM) (species and communities) with GeoAI in Python.
Description: In this module is considered integrated and correlative SDMs integrating species occurrence datasets, climatic and other environmental predictor variables, modelling algorithms, evaluation methods and widely Python packages used for SDM AI studies. The influence of ecological drivers (e.g. climate, soil, geology) and the possible erosion or expansion of their envelopes of suitability is required by decision makers in order to detect scalable indicators in Natura 2000 sites, to predict Habitat Suitability (HS) under climate change for the dominant species, endemic, rare, invasive, Threatened and Vulnerable (IUCN) species; and also to detect environmental impacts in Concessioned Mining Areas.
Table of contents:
I. Introduction to the Species Distribution Modeling (SDM) with Artificial Intelligence (AI) (“SDM AI”) module
i) What is integrated and correlative Species Distribution Modeling (SDM).
ii) Data sources and software used in the course.
iii) Introduction to Python for habitat suitability modelling.
II. The Basics of GeoAI for Species Distribution Models (SDMs).
i) Raster data for ingestion in SDM AI models (integrated and correlative).
ii) Accessing GBIF Data, Crowdsourcing Data and Copernicus Land Monitoring Services.
iii) Additional sources of Species Geo-location Data.
iv) Extraction Species Geo-location Data from Geo Apps for integration in SDMs.
v) Access and ingestion of Climate, Elevation & derived morphometric variables, Earth Observation Data (EOD) spectral index & other Essential Biodiversity Variables (EBVs) via Python.
III. Data fusion of Spatial Data for SDMs.
i) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the Presence & Absence Data and environmental variables.
ii) Region of Interest (ROI) & Sampling design.
iii) Vectorial and Raster operations to a given ROI.
iv) Spatial downscaling and upscaling.
v) Data visualization techniques.
IV. SDM Techniques for Habitat Suitability (HS).
i) Exploratory analysis – Canonical-Correlation Analysis (CCA), Principal Component Analysis (PCA) and correlation.
ii) Presence and pseudo-absence datasets.
iii) Application of predictive modelling techniques (Bioclim, Maximum Entropy (Maxent), Environmental Distance (Domain)).
iv) Model evaluation.
V. Artificial Intelligence (AI) algorithmization for SDM.
i) SDM using deep learning in Python.
ii) Generalised Linear Model (GLMs).
iii) Support Vector Machines (SVM).
iv) Artificial Neural Networks (ANN).
v) Random Forest (RF).
vi) Gradient Boosting Machine (GBM).
vii) Genetic Algorithm for Rule-set Production (GARP).
viii) Ecological Niche Factor Analysis (ENFA).
ix) Model Evaluation.
Software:
– Jupyter Notebook: Python (eg. package sdmdl).
– QGIS with plugins & Python scripts.
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 2 – Surface motion monitoring (InSAR) & DEM comparison analysis with GeoAI in Python.
Description: Interferometric Synthetic Aperture Radar (InSAR) is a powerful technique for the estimation of multi-temporal surface deformation processes, allowing a detailed monitoring of ground and structures through the use of sensors installed on SAR satellites (Sentinel-1, TerraSAR-X). In this module you will investigate the Mining Subsidence Based on Multi-Temporal InSAR and Time-Series Analysis of the Small Baseline Subset for different category mines. Interferometric Synthetic Aperture Radar (InSAR) is an effective way to measure changes in land surface altitude. The InSAR images should cover the entire area of surface subsidence. InSAR is used to determine the limits of deformation as a result of caving, to a level of 1 mm over one year. In the field, this measurement of small deformation extends beyond the visible crack limit.
Table of contents:
I. Introduction to Interferometric Synthetic Aperture Radar (InSAR) analysis.
i) Concepts and fundamentals of TOPS Interferometry.
ii) Process interferograms using the SNAP ESA – Sentinel-1 Toolbox
TOPSAR Interferometry
iii) How to distinguish between different sources of noise in InSAR data and apply appropriate corrections.
iv) How to interpret interferograms and prepare them for modeling.
v) How to use pixel offset tracking to measure large surface motions such as those related to mining exploration and environmental monitoring.
vi) How to use Python The Miami InSARTime-series software in Python (MintPy) to perform time series analysis on coregistered stacks of InSAR data compiled using SNAP ESA.
II. Data fusion of Spatial Data for InSAR analysis.
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Coregistration.
vii) Data visualization techniques.
III. Interferogram Formation and Coherence Estimation.
i) Coherence Estimation.
ii) Goldstein Phase Filtering.
iii) Phase Unwrapping.
iv) Phase to Displacement.
v) Terrain Correction.
IV. Use cases
i) Application of InSAR analysis in Open Pit Mines (Romania and Portugal).
ii) Application of InSAR analysis in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package The Miami InSARTime-series software in Python (MintPy)).
– SNAP ESA – Sentinel-1 Toolbox TOPSAR Interferometry
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 3 – Geosimulation of LULC Changes and Landscape metrics with GeoAI in Python.
Description: This GeoSDM module aims to develop geosimulation hybrid models in different category mines. The geosimulation (geocomputation) is an emergent field of analysis embracing heuristic search, artificial neural networks and cellular automata, Markov chains, among others.
Table of contents:
I. Introduction to Land Use/land Cover Changes and Prediction of Future Changes.
i) Concepts and workflows of Geosimulation.
ii) Cellular Automata and Markov-Chains.
iii) Prediction of the future land use and land cover of Concessioned Mining Areas.
II. Data fusion of Spatial Data for LULC mapping models.
i) Multispectral time series database for training LULC mapping models with Artificial Intelligence (AI).
ii) Coordinate Reference System (CRS), Open Geospatial Consortium (OGC) standards of the LULC products (CLC, COS, LUCAS survey).
iii) Standardization and harmonization of LULC legends.
iv) Region of Interest (ROI).
v) Vectorial and Raster operations to a given ROI.
vi) Combining products across time and space.
vii) Re-sampling and selection of agreement threshold.
viii) Data visualization techniques.
III. Detection of the changes in the LULC in Python.
i) Image classification.
ii) Change detection analysis.
iii) LULC modeling and future scenarios.
iv) Transition analysis and validation of LULC simulation model.
v) LULC classification and accuracy assessment.
IV. Use cases.
i) Application of geosimulation in Open Pit Mines (Romania and Portugal).
ii) Application of geosimulation in Underground mines (Finland).
Software:
– Jupyter Notebook: Python (eg. package Cellular Automaton in Python (PyPi)).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
GeoSDM module 4– Smart Flood detection with GeoAI in Python
Description: In this module, you will validate that the GeoAI solution to obtain a rapid flood mapping product, showing promising results which are comparable to the results produced by the trained human analysts. Smart FLOOD AI-based technology delivers Large-scale and High resolution Flood Hazard Maps at lower cost respect the standard hydrodynamic modeling approaches. SmartFLOOD integrates large scale and high resolution open data and artificial intelligence algorithms for supporting robust evidence-based decision-making in flood risk management. The SmartFLOOD’s consist in an extrapolation to wider areas and a downscaling to higher spatial resolution.
Table of contents:
I. Introduction to hydrological modeling system for flood forecasting.
i) Concepts and fundamentals of Flood detection with Artificial Intelligence (AI) to identify inundation areas during riverine floods.
ii) Implement a GeoAI workflow for disaster management solutions.
iii) Smart FLOOD AI-based technology to deliver Large-scale and High resolution Flood Hazard Maps.
II. Data fusion of Spatial Data for Flood Detection
i) Sentinel-1 Interferometric Wide Swath Products.
ii) Coordinate Reference System (CRS) and Open Geospatial Consortium (OGC) standards of Sentinel-1.
iii) Definition of Region of Interest (ROI).
iv) Preparation and Download the Data.
v) Vectorial and Raster operations to a given ROI.
vi) Data visualization techniques.
III. Flood Forecasting and Impact Assessment.
i) Artificial Intelligence (AI) algorithmization for flood detection.
ii) Deploy AI models for near real-time analysis.
iii) Utilize deep learning-based model inference to detect and respond to flood events.
iv) Clustering UML and Artificial Neural Networks (ANN) in Python to detect flood zones.
v) Future scenarios and economic damages in QGIS (plugin Floodrisk).
vi) Validation of Flood simulation model with in-situ data and drone sensing (eg. NOAA).
vii) Flood hazard map classification and accuracy assessment.
IV. Use case.
i) Developing a flood hazard map for a selected river basin in Houston (Texas, USA) (after Hurricane Harvey).
ii) Improving the end-to-end flood early warning system in Houston (USA).
Software:
– Jupyter Notebook: Python (eg. package Keras).
– QGIS (plugin Floodrisk).
Workload: 20 hours of online classes, approximately 14 hours of individual home assignments, 7 hours of monitoring and assignment questions.
Duration: 5 weeks
Dates: Tuesdays and Thursdays 4 p.m. – 6 p.m. Lisbon/London time, starting on May 21st 2024. Assignments’ support optional meetings – Fridays 4.p.m.
Trainer: Francisco Gutierres. He has PhD in Physical Geography (with specialisation in SDM models and Landscape metrics at local scales – Natura 2000 sites in the West Coast Portugal), Master in Conservation Biology, Postgraduation in GIS Applied to Sciences of the Earth, PhD specialisation in Geo-Ecological Data Analysis and a Licenciatura’s Degree in Biophysical Engineering – Environmental Planning and Management. Before joining the OPT/NET B.V as Geomatics Manager, Francisco worked for 20 years as a GIS and Environmental specialist in national and local government and private industry.
Related subject
- 75In the business world, market analysis, and urban planning, understanding influence areas is crucial for making informed and strategic decisions. These areas represent the geographical extent where a specific activity, service, or location has a significant impact. Let's explore the methods for calculating and delineating these areas to gain valuable…
- 72In a world where technology intertwines with our daily lives, it is fascinating to witness how Geographic Information Systems (GIS) have revolutionized the way we approach archaeology and heritage management. Archaeology and GIS: A Powerful Alliance 🧭🏺 Archaeology, which relied on traditional methods for centuries, now embraces the digital era…
- 69In the digital age we live in, technology plays a crucial role in the evolution of various industries, and the real estate market is no exception. One of the most fascinating innovations gaining prominence is the application of remote sensing, using satellites and drones, to revolutionize how we understand,…