Quality assurance of near-surface wind velocity measurements in Mexico

Abstract

A comprehensive quality assurance (QA) procedure applied to the wind velocity records of the automated surface observation network of the Mexican National Weather Service has been conducted for the first time. The analysis covers a total of 139 stations with site-specific measurement periods ranging from 1 to 12 years. The QA process integrates 11 QA checks grouped into 4 categories (plausible value check, internal consistency, temporal consistency, and systematic errors). The fraction of the records flagged by the quality-control (QC) filters is substantial, with nearly 26% of all wind speed records and 5% of all wind direction records showing anomalies. Most error candidates were found to fall in the systematic error and temporal consistency categories, with 10 and 12% of the cases occurring in these categories, respectively. A significant variation of data quality over time was observed, showing both a build-up of errors over time in early stations as well as the results of a learning curve in stations installed more recently. In the case of the earliest stations, installed between 1999 and 2001, a peak in anomalies is observed about 4 years after installation with a subsequent reduction and a second peak after about 10 years. While stations installed between 2002 and 2004 show similar patterns, with error frequency peaking after about 5 years, stations installed more recently, between 2002 and 2006 show a greatly improved behaviour. The results communicated in this paper should provide useful input information for validation of mesoscale wind maps and improved data assimilation procedures.