The quality control of weather data is a necessity and a responsibility of meteorological services that store, distribute, and use these data. In the present work, a newly designed quality control procedure for daily rainfall data is presented after it has been adjusted and tested with more than 10^7 data from 1726 daily rainfall measurement sites in Catalonia. It is applicable to data from different origins (e.g., automatic weather stations or manual historical measurements). The procedure is focused on relative comparison of daily data with reference stations that are automatically selected after an initial estimation of their quality and a proximity study regarding location and correlation. The presented procedure has been verified taking advantage of an available network in the study area that has been routinely quality controlled by technicians of the Meteorological Service of Catalonia. The newly designed quality control procedure for daily precipitation yields good results, especially for extreme values: type I error under 10% is found for values up to 150 mm (error decreasing for lower values) and type II error is under 16% when reported values are twice a measure of 50 mm or more (error decreasing for more extreme values). After the application of the quality control procedure, a selection of series with the minimum desired quality is achieved.

Peer Reviewed