We are studying optical and Synthetic Aperture Radar (SAR) applications for crop monitoring and production forecasting for both large scale agricultural systems and smallholder systems at the field to national scales. This involves developing and refining models for yield forecasting, cropland and crop type mapping, and crop condition assessments.
Optical satellite data have been widely used for monitoring crop condition and productivity. However, the presence of clouds impedes data collection and as such, these cloudy conditions are a significant impediment for monitoring crop development operationally. The longer wavelengths propagated by SAR sensors are unaffected by clouds. Consequently, the use of SAR data for mapping and monitoring agricultural landscapes helps to consistently deliver crop maps. We are using a combination of Sentinel-1 and -2 and Landsat-8 to generate crop maps at 20m spatial resolution at the field to national scales. Our previous studies demonstrated that SAR signals are highly sensitive to crop structures and promising results were derived for both crop type mapping and for estimations of crop productivity indicators such as crop leaf area index (LAI) and biomass. In this project, we are using a combination of SAR and optical data for crop yield estimation and crop type mapping. There are two objectives for this data combination; to generate a high time series of crop maps regardless of weather condition and to improve the crop map accuracies by integrating input features from SAR and optical data. Machine learning algorithms are trained and tested to find the best optical and SAR features for crop mapping.
The upcoming NISAR satellite will provide multi-frequency SAR data. This multi-frequency and multi-polarization satellite is sure to provide very interesting SAR data for agricultural applications. Therefore, the developed models in this project will be adapted for multi-frequency SAR data upon the availability of the NISAR data. Given that the NISAR satellite has two frequencies, S-band and L-band, it is expected that using its data will significantly improve the accuracy of crop maps. As our contribution to the Group on Earth Observations Global Agricultural Monitoring Initiative (GEOGLAM), the final model/models developed in this project are delivered as a cloud-based tool. They are implemented operationally at national and international scales and capacities are developed for our national and international partners to use the developed products.
Hosseini M., Becker-Reshef I., Justice C., 2020, Crop Harvest Monitoring Using Polarimetric SAR Parameters, IGARSS 2020.
Hosseini M., McNairn H., Mitchell S., Davidson A., Dingle Robertson L., 2019, Synthetic Aperture Radar and Optical Satellite Data for Estimating the Biomass of Corn, International Journal of Earth Observation and Geoinformation, Vol 83, 101933.
Hosseini M., McNairn H., 2017, Using multi-polarization C- and L-band synthetic aperture radar to estimate wheat fields biomass and soil moisture, International Journal of Earth Observation and Geoinformation, 58, 50-64.