NASA Harvest, NASA’s Food Security and Agriculture Program, is a global consortium with contributions from people of many different backgrounds, specialties, and interests. What unites us all is a dedication to bolstering food security around the world through Earth observation applications, and a shared passion for technology that improves lives. We are proud of the work that we do and the people who produce it. This feature introduces the people of NASA Harvest, showcasing the members of our organization and how their efforts support a food-secure future.
Dr. Hosseini is an Associate Research Professor of the NASA Harvest Hub in the Department of Geographical Sciences at the University of Maryland, College Park. His research is related to optical and SAR applications to crop monitoring and production forecasting for both large scale agricultural systems and smallholder systems at the field to national scales. His research involves developing and refining models for yield forecasting, cropland and crop type mapping, and crop condition assessments.
From 2017-2019, Dr. Hosseini was a co-lead of an international project involving 18 countries, studying the best practices for crop biomass and leaf area index (LAI) estimations and crop type classification using Synthetic Aperture Radar (SAR). As part of this project, he helped develope global models for monitoring crop conditions over five globally important crop types (corn, wheat, soybeans, canola and rice).
How did you first get interested in remote sensing and agricultural monitoring?
I first got interested in remote sensing when I took an undergraduate course on different types of satellites, their specifications, and how their data was used in real world applications. I enjoyed the class so much that I decided to pursue a Masters in this field and focused my thesis on geometric correction of satellite data. For my PhD, my research was more operational and I investigated soil moisture estimation modeling using a combination of Synthetic Aperture Radar (SAR) and optical satellites. During the course of my PhD research, I was introduced to the agricultural applications of remote sensing and how satellite imagery enables the monitoring of agricultural fields over wide regions where ground measurements are very challenging and expensive. I found this area of research interesting and I continued working on it through my Postdocs and in my current role at NASA Harvest.
Why is agricultural monitoring important? What are the benefits remote sensing has on things like agricultural production, food security, and market stability?
It is estimated that the world population will grow to 10 billion by 2050. This increase will raise demand for agricultural production and likely increase global food insecurity Advanced technologies like remote sensing help us monitor agricultural fields at a global scale so we can detect issues on the crops such as water stress or disease and act quickly to minimize damage and maximize production. Remotely sensed data also allows us to make better management decisions by monitoring soil moisture over cropland and analyzing the production impacts of irrigation. In terms of markets, we’re developing increasingly accurate models to forecast crop yield and global production. These models allow us to predict output and identify areas of concern which creates more transparent and resilient markets.
Before coming to the University of Maryland and NASA Harvest, you co-lead several projects for JECAM in Canada. What was the focus of some of the projects that you lead?
In the JECAM project, I was a co-lead of a SAR inter-comparison project with the goal of testing and comparing all the available models that are used by our international partners around the world for crop condition and crop type detection. I led the crop condition component including crop leaf area index (LAI) and crop biomass, and my colleague at Agriculture and Agri-Food Canada was the lead for the crop type. We had more than 20 international partners which collaborated and provided ground data. This project produced LAI and biomass SAR-based estimation models for five major commodity crops including wheat, corn, soybeans, rice and canola.
You mentioned that you used SAR data to measure LAI and biomass. There has been an increased interest in the use of SAR data over the past several years. A lot of your research includes radar data, particularly synthetic aperture radar (SAR) imagery. What made you first get involved with SAR and why do you think it's important to expand our use of it?
I’ve been working with SAR since I started working on my PhD. My doctoral research focused on soil moisture estimation modeling over vegetated areas using SAR data. SAR is a really useful tool as its high wavelengths allow it to penetrate cloud cover and so, unlike optical satellites, SAR sensors are more operational during cloudy days and particularly in areas with high cloud cover like tropical regions. Additionally, as SAR satellites are active (meaning they have their own source of energy), they are able to collect data during the day and night unlike optical satellites which are reliant on reflected sunlight. The longer wavelengths of SAR signals mean that they can also penetrate crop and tree canopy as well as the soil depending on the specific sensor’s wavelengths and the current crop growth stage. SAR signals are very sensitive to the dielectric constant of soil and this sensitivity allows us to measure variables like soil moisture or vegetation water content based on how the signal reflectance varies. Another interesting application of SAR data is that its sensitivity to structure of the crops makes it useful for monitoring crop growth throughout the growing season. This incredible utility has led to a number of private companies increasingly investing in the development of SAR satellites and applications over the past decade. I’m excited for this private and public sector interest and capacity-building and am looking forward to how else we can expand SAR.
Several of your recently published articles have focused on measuring soil moisture with radar EO data. Can you talk a little about why accurate soil moisture measurements are important for agricultural research?
One of the most important factors for crop productivity is soil moisture. Accurate measurement of soil moisture also allows farmers to understand the potential need for and effectiveness of their irrigation scheme. Irrigation is expensive for farmers and it can have varying levels of success depending on different factors. Being able to evaluate the effectiveness of different irrigation management schemes can help farmers make more economic decisions for future harvests. As climate change increasingly impacts weather patterns like precipitation and temperature and makes drought more likely in certain regions, fast and accurate analysis of soil moisture conditions will become even more vital.
You mentioned above that one of the benefits of radar over optical data is the former’s ability to penetrate cloud cover. You recently co-authored a study on how radar can also penetrate crop canopy to get soil moisture measurements. This allows for measurements of soil moisture in further along stages of the growth cycle. However accurate measurements require the user to account for crop canopy interference in reflectance values. Can you talk a little about how you accounted for crop canopy to acquire accurate soil moisture measurements?
Being able to monitor soil moisture throughout a crop’s growing cycle is really important for understanding crop conditions. Some higher wavelength SAR data from L-band and P-band satellites are able to penetrate vegetation to the soil below while shorter wavelength SAR satellites like C-band, X-band, and K-band can be obscured. One of the most widely used sources of SAR data is the European Space Agency’s C-band Sentinel-1 mission. Sentinel-1 data is free to access and has high spatial and temporal resolution, making it very useful for agricultural monitoring. However, as it is a C-band satellite, it is limited in its penetration capabilities. Our study looked at how Sentinel-1 data could be operationalized to pierce crop canopy and measure soil moisture throughout the entire growing season.
In basic terms, SAR-based soil moisture measurements are calculated using the amount of backscatter, or reflectance, of the SAR pulse recorded by the satellite sensor as it passes over the Earth’s surface. The amount of backscatter varies as the moisture content in soil changes. Likewise, the amount of vegetation influences the amount of backscatter recorded. In our study, we used a technique called polarimetric decomposition to “decompose” or separate the backscatter caused by vegetation and the backscatter caused by soil moisture. This allowed us to account for and exclude the backscatter attributed to vegetation and measure only the backscatter caused by the moisture content of the soil.
What was the result of your study? What do your results mean for potential future agricultural applications of the Sentinel-1 platform?
Our study showed really promising results and we were able to get soil moisture measurements through crop canopy with high accuracy. While we’ve shown it can be done, our study was limited to maize crops grown at a particular growth stage. To increase the transferability of the technique, more research is necessary on using our method on different kinds of crops at various growth stages to ensure that our method is robust in as many environments as possible. As our method is tested in more and more conditions, the applicability of the Sentinel-1 platform will increase and provide greater opportunities for SAR-based soil moisture monitoring.