Use of Landsat imagery to assess spatial variability in the normalized difference vegetation index in the contributing area of USGS groundwater wells to help describe land-use changes across the Fort Peck Indian Reservation, Montana, 1990-2020

This exploratory study examined vegetation changes in agricultural fields using imagery analysis in the surface areas overlying the Flaxville aquifer and other aquifers in Montana from 1990 to 2020. Agricultural fields were manually digitized and can be found in flaxville_ag_fields.shp. This analysis used surface reflectance data from Landsat satellites to calculate the 95th percentile NDVI (Normalized Difference Vegetation Index) annually for each pixel. To assess spatial and temporal variability, a single image with 31 bands, each representing the 95th percentile NDVI value of a year, was compiled. Spatial variability was assessed by calculating the mean 95th percentile NDVI for digitized fields and finding the zonal mean of absolute differences within each field. Contributing areas for 90 USGS wells were delineated, and the weighted mean variance of NDVI values within these areas was calculated to investigate temporal variability. Linear regressions were performed on these variances. The linear regression results are listed in flaxville_weighted_linear_regressions.csv.