During Earth Week at Goddard, Oct. 11-14, 2022, five members of the Goddard community gave virtual presentations. On October 11^th, GESTAR II member Akiko Elders (618/MSU) presented on the use of Earth observing systems, remote sensing, and machine learning models to address and combat food insecurity due to crop conditions, particularly in Burkina Faso, located in West Africa. She provided the following summary:

"Earth observing systems provide a unique opportunity to monitor and evaluate crop conditions globally. This is particularly important for evaluating regions with food insecurity. This project used Sentinel-2 and Machine Learning to estimate crop type and yield in Burkina Faso with greater than 80% accuracy in rainy season crop type and rice yield estimates. A unique aspect of this work is the use of Machine Learning models to make uncalibrated, interannual predictions of dry season crop type, which is critical for evaluating food availability in data scarce regions with limited agricultural data. The uncalibrated models estimated subsequent dry season crop type with as much as 60% accuracy, indicating some utility in uncalibrated estimates of crop type for years without training data. Ultimately, the ability to use affordable and timely remote sensing-based methods is essential for remote agricultural monitoring and evaluation and appropriate humanitarian interventions to combat food insecurity."

During your presentation, you mentioned working with Millennium Challenge Corporation. Would you expound on that?

"The work we are doing with MCC is part of an interagency agreement where NASA provides remote sensing and Machine Learning expertise. MCC is interested in how remote sensing can be used to assess their Agricultural Development Projects (ADP). This is a four institution collaboration working in partnership towards this goal."

Who are the end-users of your findings, and how would this be applied?

"Overall, the information from our work is to benefit MCC's mission. They invest in ADPs such as building irrigation infrastructure to help small holder farmers in developing nations transition from subsistence to small-scale commercial farming in efforts to combat food insecurity. Traditionally, they evaluate their ADPs with field work, which is expensive and can be limited by social conflicts. They are looking to us to explore how they can accurately understand the growth of their projects in years when they cannot collect data. MCC has a number of ADPs across Africa, so our work with them will continue as they expand to new perimeters. We will leverage what we have learned in Burkina Faso and follow the development of a new perimeter, in West Africa, over time."

In August 2022, Dr. Elders published an article on the work in Burkina Faso: Elders, A., Carroll, M.L., Neigh, C.S.R., D'Agostino, A.L., Ksoll, C., Wooten, M.R., Brown, M.E., 2022. Estimating crop type and yield of small holder fields in Burkina Faso using multi-day Sentinel-2. Remote Sensing Applications: Society and Environment 27, 100820. https://doi.org/10.1016/j.rsase.2022.100820.

Not only is Dr. Elders a postdoctoral assistant research scientist in Code 618, the Biospheric Sciences Lab, she also holds the same position in Code 615, the Cryospheric Sciences Lab at NASA Goddard. There, she currently "is working on a project that evaluates sea ice representation in NASA GISS ModelE. Climate models have a lot of variability in how they represent Arctic climate, such as fluxes between the ice and atmosphere. We are reviewing how the sea ice is represented and how that impacts the Arctic climate and ways to potentially improve the representation of the sea ice and climate."