HadISD: a quality-controlled global synoptic report database for selected variables at long-term stations from 1973–2011

摘要

This paper describes the creation of HadISD: an automaticallyquality-controlled synoptic resolution dataset of temperature,dewpoint temperature, sea-level pressure, wind speed, wind directionand cloud cover from global weather stations for 1973–2011. The fulldataset consists of over 6000 stations, with 3427 long-term stationsdeemed to have sufficient sampling and quality for climateapplications requiring sub-daily resolution. As with other surfacedatasets, coverage is heavily skewed towards Northern Hemispheremid-latitudes.The dataset is constructed from a large pre-existing ASCII flatfiledata bank that represents over a decade of substantial effort at dataretrieval, reformatting and provision. These raw data have hadvarying levels of quality control applied to them by individual dataproviders. The work proceeded in several steps: merging stations withmultiple reporting identifiers; reformatting to netCDF; qualitycontrol; and then filtering to form a final dataset. Particularattention has been paid to maintaining true extreme values wherepossible within an automated, objective process. Detailed validationhas been performed on a subset of global stations and also on UK datausing known extreme events to help finalise the QC tests. Furthervalidation was performed on a selection of extreme events world-wide(Hurricane Katrina in 2005, the cold snap in Alaska in 1989 and heatwaves in SE Australia in 2009). Some very initial analyses areperformed to illustrate some of the types of problems to which thefinal data could be applied. Although the filtering has removed thepoorest station records, no attempt has been made to homogenise thedata thus far, due to the complexity of retaining the truedistribution of high-resolution data when applying adjustments. Hencenon-climatic, time-varying errors may still exist in many of theindividual station records and care is needed in inferring long-termtrends from these data.This dataset will allow the study of high frequency variations oftemperature, pressure and humidity on a global basis over the lastfour decades. Both individual extremes and the overall population ofextreme events could be investigated in detail to allow for comparisonwith past and projected climate. A version-control system has beenconstructed for this dataset to allow for the clear documentation ofany updates and corrections in the future.