Skip to main content

Improved Satellite Technology and Radar Products Advance Cropland Mapping

SAR image

In this era of rapidly advancing technology where there is a renewed focus on increased and sustainable food production, agricultural experts are constantly working towards improved crop monitoring using the best available tools. One recent and particularly exciting development for researchers in this space is the establishment of synthetic aperture radar (SAR) as a key player for enhancing remotely-sensed agricultural monitoring capabilities. Building upon the traditional methods of using optical satellite data to track crop production “Cropland mapping with L-band UAVSAR and development of NISAR products” evaluates the usefulness of new radar products to supplement long-established agricultural monitoring methods. NASA Harvest partners and colleagues across the agriculture sector including: Applied Geosolutions, the USDA Agricultural Research Service, the University of Arkansas, NASA’s Jet Propulsion Laboratory, the University of Massachusetts, and the University of Maryland collaborated to evaluate a suite of methods for utilizing L-band (24cm) SAR data for crop monitoring and support the development of prototype NISAR Level-2 agricultural products.


Agriculture and food production impact many aspects of human livelihood, from basic nutritional needs to global food markets and governmental policy implementation. In order to evaluate food availability and forecast probable production at local, national, and global scales, it is vital to understand the type of crops produced in each specific area of interest. While NASA has supported the use of Earth observations and remotely sensed satellite data since the very early development of this technology, research and applications have typically been focused on optical data while SAR information took a temporary backseat. 


However, through the close collaboration of international space agencies, agricultural experts now have the added benefit of open access to radar data streams which was first made possible by the 2014 launch of the European Space Agency’s Sentinel-1 satellite mission. Looking to the future, the 2022 launch of NASA’s NISAR (NASA-ISRO SAR) Mission is expected to provide even more expansive access to remotely sensed radar data. This is especially intriguing for agricultural scientists because SAR instruments differ from optical instruments in a few key ways, most notably in that they allow for all-weather crop assessments and are more sensitive to crop and field characteristics (i.e., dielectric constant, roughness, orientation) than traditional apparatuses.


SAR image


For this particular study, the authors chose four “agricultural production hotspot” sites based on their landscape composition, ongoing activities, local partnerships, and flight operations. At each site, they collected crop type observations during the growing season in order to ground truth, calibrate, and validate machine learning algorithms based on UAVSAR and NISAR simulated observations which were analyzed to better understand the advantages, limitations, meanings, and performance of these data and models. They performed hundreds of classification experiments to assess scattering mechanisms, the performance of L-band SAR parameters, and the robustness of techniques for cropland monitoring. 


methods structure


While the results varied with each classification experiment, it became evident throughout the study that there is added value in time series SAR and critical temporal information with increased density, outweighing individual scattering mechanisms or machine learning classification techniques. The authors also note that as more SAR products become available, stakeholders can better “leverage physical contributions of different wavelengths and polarizations along with growing open access time series for efficient and meaningful agricultural products,” highlighting the fact that open access to as much data as possible will ultimately benefit the entire user community.


Read the full publication and explore the detailed findings in Remote Sensing of Environment.

News Date
Jan 7, 2021
Mary Mitkish, Sergii Skakun