Spatial Variation of Natural Storms

An important limitation to understand and quantify is the spatial variation of rainfall depth over a watershed.  In fact the rainfall depths, much like the radar predictions of rainfall you see on weather maps, not only varies across a large watershed, but can be substantially different over just a short distance.  While a continuous (different at every location) map of rainfall depth provides the true representation, as of yet we do not have the capability to accurately measure this on the ground.

The radar rainfall estimates you see on weather maps would appear to be the perfect solution, but they are only estimates of where the rainfall is more intense.  While they provide a good picture of where it might be raining harder relative to other locations, they actually do a poor job of predicting absolute values of rainfall depth on the ground.  This is due in part to the fact that they measure reflectance at a high altitudes.  Some of what is measured never reaches the ground, or it may be blown as it falls making the location on the ground different than what is measured.  Radar rainfall data, when adjusted by measured rain gages can be useful, but this process is still uncertain (because of the lack of gages) and not yet available in real time.  It is certain that in the future the ability to use radar data will improve, but for now most models will continue to be developed with basin average, isohyetal, Thiessen-weighted, etc. estimates.