Quantifying Uncertainty in Area and Effect Size Estimation from Remote Sensing Maps
Kerri Lu, Massachusetts Institute of Technology
Remote sensing map products are used to obtain estimates of environmental quantities. However, the quality of map products varies, and --- because maps are outputs of complex machine learning algorithms that take in a variety of remotely sensed variables as inputs --- errors are difficult to characterize. Without capturing the biases that may be present, naive calculations of population-level estimates from such maps are statistically invalid. We compare several uncertainty quantification methods --- stratification, Olofsson area estimation method, and prediction-powered inference --- that combine a small amount of randomly sampled ground truth data with large-scale remote sensing map products in order to generate statistically valid estimates. Applying these methods across five remote sensing use cases in area and effect size estimation, we find that they result in estimates that are more reliable than naive imputation (using only the map product) and generally have lower uncertainty than classical estimates (using only the ground truth). Prediction-powered inference uses ground truth data to correct for bias in the map product estimate and (unlike stratification) does not require us to choose a map product before sampling. This is the first work to (1) apply prediction-powered inference to remote sensing estimation tasks, and (2) perform uncertainty quantification on remote sensing regression coefficients without assumptions on the structure of map product errors.