What does the term "Root-Mean Square error" refer to in geospatial analysis?

The term "Root-Mean Square error" (RMSE) is indeed a statistical measure that reflects the accuracy between two datasets, making this option the correct choice. RMSE quantifies how much predicted or estimated values deviate from the observed values. In geospatial analysis, RMSE is particularly useful in assessing the accuracy of spatial data sets, such as comparing the position of features in a predicted model against their known positions in a reference dataset.

This measure is computed by taking the square root of the average of the squared differences between corresponding values. A lower RMSE value indicates better accuracy, and it helps analysts to evaluate the performance of spatial models or the quality of map outputs.

The other options relate to different aspects of geospatial analysis, but they do not adequately describe what RMSE represents. Ground truthing, for example, is a process to validate the accuracy of remote sensing data by comparing it with real-world observations, which is distinct from the statistical measure of RMSE. Comparing polygons in land cover classification involves the analysis of vector data rather than the numerical error calculations of RMSE. Digitizing features from aerial images refers to the creation of digital representations from visual sources, which is a different activity from using RMSE to assess accuracy.