Studying Crop Yield Constraints From Space
Increasing food production to meet growing demand, while at the same time preserving natural ecosystems, will require further increases in crop yields (tons of grain per hectare of land). Through breeding, and through the development of new fertilizers, irrigation systems, pesticides, and other agronomic changes, substantial increases in the maximum potential yields of our main food crops have been achieved in the past 50 years. Yet further increases will require new innovative solutions.
Although many casual observers might say that all fields look the same, for example when driving through the Central United States, in fact yields often vary by a factor of two or more between nearby fields. The gap between the potential yields of these crops and the yields actually achieved represents food we could be producing but aren’t, stomachs that could be filled but instead go empty. The focus of this project is to determine what factors are contributing to this yield gap in selected regions around the world, and at what magnitude, so that we can better understand where to direct future research aimed at closing that gap.
The example above shows estimates of wheat yields derived from satellite images of the Indian Punjab, one of the most intensive and important wheat growing regions in the world. The first image shows estimate for a six district area covering over 1 million hectares, while the second shows a closer look at the full resolution (28.5 m) of the data. Red indicates fields with yields near 5 ton/ha, while blue indicates yields of roughly half this value. We can use these maps for several years, combined with data on climate, soil type, and geographic and manmade features such as waterways and roads, to increase our understanding of the factors affecting grain yield in the Punjab. These techniques are transferable to other regions of the world, so in time we may be able to diagnose yield gaps across the globe, leveraging the power of remote sensing to help produce more food and potentially use less water and other inputs in the process.