Precipitation measurement in upland environments – problems, progress and prospects

Abstract

Efforts to measure precipitation extend back into history for more than 500 years, and to date remain challenging owing to the effects of wind, snow and local variability in precipitation-generating processes. Errors in precipitation amounts can easily exceed 20%, and prominent inconsistencies are known where catchment runoff appears to exceed precipitation inputs, even without allowing for evaporative losses. One solution to the unreliability of precipitation data is to simply avoid precipitation estimates altogether and use runoff data instead, but this does a disservice to scientists who may wish to see closure of the water balance, may need to undertake design studies (e.g. requiring probable maximum flood estimates) or to better understand hydrological processes in a changing environment. The presentation reviews insights gained from analysis of 100+ years of rainfall data and evaluates the prospects for producing rainfall observations which are fit for purpose.