Hydrological models combining physics with AI are becoming popular for predicting river flow, but are often unnecessarily complex. We tested these models across US river basins and found three key results: simpler designs perform equally well, extensive input data adds little value, and time-varying parameters do not represent actual physical processes. These results challenge assumptions that complexity improves predictions or understanding, arguing instead for simpler hybrid model development.

Long-term overuse of groundwater has lowered water levels and increased pollution in parts of northern China. We studied whether adding river water back into the ground can restore groundwater and reduce nitrate pollution. Using computer simulations based on field data, we found that recharge can significantly raise water levels and lower nitrate levels, mainly by dilution rather than natural removal processes.