We may observe that the availability of remote sensing data is growing. This, along with the advancement and availability of computer tools, has enabled even the most casual map producer to generate prediction maps.
However, such maps should always be seen in light of the inherent uncertainty, as this will give the maps more significance. Moreover, prediction maps are often intermediary products from which the ultimate output is further derivable or may serve as the basis for a decision; hence, it is still important to view them with the related uncertainty.
But it is impossible to stop people from creating such maps. Thus, the burden transfers to the map readers to evaluate the maps as they use them. For the map reader to be able to estimate the prediction variability, the map producer must supply the required metadata. One example of metadata is the variance-covariance matrix of the estimates (for the parametric regression models). If this information is missing, the published map might not be very useful.
References:
Fassnacht, F.E., White, J.C., Wulder, M.A. and Næsset, E., 2024. Remote sensing in forestry: current challenges, considerations and directions. Forestry: An International Journal of Forest Research, 97(1), pp.11-37.
Whelan, B.M. and McBratney, A.B., 1999, January. Prediction uncertainty and implications for digital map resolution. In Proceedings of the Fourth International Conference on Precision Agriculture (pp. 1185-1196). Madison, WI, USA: American Society of Agronomy, Crop Science Society of America, Soil Science Society of America.
