Researchers in Nepal used a new method called DSMART to create a more detailed soil map by breaking down an existing map. They used data from remote sensing, climate, and other sources to predict soil types. The new map showed more detailed soil information than the original one. The accuracy of the predictions was around 40% for overall soil types and 22% for specific soil groups. Factors like geology, temperature, elevation, valley depth, and land cover were important in predicting soil types. This new detailed soil map can be helpful for research, planning, and managing land resources in Nepal.