Description
Monitoring crop growth is important for assessing food production, enabling optimal use of the landscape, and contributing to agricultural policy. Remote sensing methods based on optical and/or radar sensors have become an important means of extracting information related to crops. Optical data is related to the chemical properties of the vegetation, while radar data is related to vegetation structure and moisture. Radar can also image the Earth’s surface regardless of almost any type of weather condition.
This three-part, advanced training builds on previous ARSET agricultural trainings. Here we present more advanced radar remote sensing techniques using polarimetry and a canopy structure dynamic model to monitor crop growth. The training will also cover how to apply machine learning methods to classify crop type using a time series of Sentinel-1 & Sentinel-2 imagery. This series will include practical exercises using the Sentinel Application Platform (SNAP) and Python code written in Python Jupyter Notebooks, a web-based interactive development environment for scientific computing and machine learning.
This webinar series is a collaboration between ARSET, Agriculture and Agri-Food Canada (AAFC), European Space Agency (ESA), University of Stirling, University of Ljubljana, and the CEOS Working Group on Capacity Building & Data Democracy (WGCapD).
This training is also available in Spanish.
- Fundamentals of Remote Sensing
- Introduction to Synthetic Aperture Radar
- Agricultural Crop Classification with Synthetic Aperture Radar and Optical Remote Sensing
- Mapping Crops and their Biophysical Characteristics with Polarimetric SAR and Optical Remote Sensing
Software Prerequisites (See Instructions Within Each Part)
- SNAP
- PolSARpro
- QGIS
- Jupyter
By the end of this training attendees will be able to:
- Monitor crop growth with polarimetric time series SAR data from Sentinel-1
- Examine crop growth using a canopy structure dynamic model and time series of Sentinel-1 imagery
- Classify crop type using a time series of radar and optical imagery (Sentinel-1 & Sentinel-2)
This webinar series is intended for local, regional, federal, and non-governmental organizations from agriculture and food security related agencies to use radar and optical remote sensing applications in the domain of agriculture for crop mapping and monitoring.
- Three, 2.5-hour sessions
- The morning session will be presented in English: 10:00 AM – 12:30 PM ET (16:00 – 18:30 CEST)
- The afternoon session will be presented in Spanish: 13:00 – 15:30 PM ET (19:00 – 21:30 CEST)
Trainers: Sean McCartney & Erika Podest
Guest Trainer: Armando Marino (University of Stirling)
- Identifying crops with polarimetric SAR (PolSAR) time series from Sentinel-1 using Random Forest and other Machine Learning algorithms in Python Jupyter Notebook
- Q&A Session
Optional for Part 1: Although not a prerequisite, in Part 1 there will be a demonstration of PolSARpro and Anaconda which will be used to identify crops with polarimetric SAR (PolSAR) time series data.
If you wish to replicate the Part 1 demonstration, please follow the software instructions found below to install Anaconda on your machine. Download times can vary depending on internet speed.
Downloading and Installing Anaconda
We will have the video of this demonstration within the training recording available within 48 hours after the presentation for you to go through at your own pace
Materials:
Trainers: Sean McCartney & Erika Podest
Guest Trainers: Krištof Oštir & Matej Račič (University of Ljubljana)
- Detecting crop type with machine learning and time series data from Sentinel-1 & Sentinel-2 imagery
- Q&A Session
Optional for Part 2: Although not a prerequisite, in Part 2 there will be a demonstration on detecting crop type with machine learning and time series data from Sentinel-1 & Sentinel-2 imagery.
If you wish to replicate the Part 2 demonstration, please create a Sentinel Hub login, send the registered email from Sentinel Hub to matej.racic@fgg.uni-lj.si and you will receive credits for processing and advance use.
Sentinel Hub Credentials instructions
Materials:
Trainers: Sean McCartney & Erika Podest
Guest Trainers: Heather McNairn, Emily Lindsay, & Xianfeng Jiao (AAFC/AAC)
- Monitoring crop growth using a canopy structure dynamic model and time series of Sentinel-1 imagery
- Q&A Session
Optional for Part 3: Although not a prerequisite, in Part 3 there will be a demonstration on monitoring crop growth using a canopy structure dynamic model and time series of Sentinel-1 imagery.
If you wish to replicate the Part 3 demonstration, please Download and install SNAP and PolSARpro.
Attendees will need access to Google Drive and Google Colab. To access these resources, users must use an email ending in ‘gmail.com’.
Follow these instructions to download the S1 SLC image.
- Sign in to Alaska Satellite Facility Vertex
- Search Type: List
- Under “List of Scene names” copy and paste the entire scene name: S1B_IW_SLC__1SDV_20190727T002316_20190727T002343_017312_0208E8_9A56
- Download the corresponding files “L1 Single Look Complex (SLC)” & “XML Metadata (SLC)”
We will have the video of this demonstration within the training recording available within 48 hours after the presentation for you to go through at your own pace
Materials:
- Presentation Slides
- Carmen_CornForRegression.csv (right click and save)
- SLC Data
- SCL metadata
- Homework (Due April 25, 2023)
- Q&A Transcript
