TITLE:
Planning Studies for Compost Application to California Rangelands Using Landsat Satellite Imagery, Carbon Modeling, and Machine Learning
AUTHORS:
Christopher Potter, John Wick
KEYWORDS:
California, Rangeland, Compost, Satellite Imagery, Soil Carbon Modeling, Machine Learning
JOURNAL NAME:
Natural Resources,
Vol.16 No.2,
February
28,
2025
ABSTRACT: Previous field measurements in rangelands throughout California have shown that spreading a relatively thin layer of compost on the soil surface of grasslands can enhance water-holding capacity and provide stabilized, slow-release nutrients to support long-term belowground carbon capture and storage. Compost-treated grasslands have been shown to consistently absorb more CO2 from the atmosphere into the plant and soil cover, more than that was being lost to microbial respiration for many years after a single organic matter application. The purpose of this new study was to optimize the long-term increase and restoration of soil carbon pools across the state of California, based on a combination of state-wide satellite image analysis, soil carbon modeling, and Machine Learning.