The Conservation Biology Institute (CBI) will make a presentation on recent work using the latest remote sensing imagery, machine learning, and the best available training data to characterize vegetation on Conservation Reserve Program (CRP) lands. A rich suite of enviro-climatic data, multiple sources of satellite imagery, and machine learning modeling techniques were deployed on the cloud-computing platform, Google Earth Engine, to maximize the accuracy of predicting land cover for study areas in Washington, Colorado, and Kansas, where CRP Conservation Reserve Program (CRP) Grasslands holdings are most prevalent.

In a parallel project focused on mapping forested CRPs in Mississippi, CBI developed predictive maps of tree height, tree density, biomass, basal area, and forest type using Random Forest modeling. Numerous satellite-derived indices from the European Space Agency’s (ESA) Sentinel-1 and Sentinel-2 sensors, in addition to soils and topography data, were used as predictor inputs. We then refined these predictive models to increase accuracy by incorporating preliminary data products derived from NASA’s spaceborne LiDAR mission - the Global Ecosystem Dynamics Investigation (GEDI).