The Quality Control of Long-Term Climatological Data Using Objective Data Analysis

Jon K. Eischeid Cooperative Institute for Research in Environmental Sciences (CIRES), University of Colorado, Boulder, Colorado

Search for other papers by Jon K. Eischeid in
Current site
Google Scholar
PubMed
Close
,
C. Bruce Baker National Climatic Data Center, NOAA, Asheville, North Carolina

Search for other papers by C. Bruce Baker in
Current site
Google Scholar
PubMed
Close
,
Thomas R. Karl National Climatic Data Center, NOAA, Asheville, North Carolina

Search for other papers by Thomas R. Karl in
Current site
Google Scholar
PubMed
Close
, and
Henry F. Diaz Climate Diagnostics Center, NOAA/ERL, Boulder, Colorado

Search for other papers by Henry F. Diaz in
Current site
Google Scholar
PubMed
Close
Full access

Abstract

One of the major concerns with detecting global climate change is the quality of the data. Climate data are extremely sensitive to errant values and outliers. Prior to analysis of these time series, it is important to remove outliers in a methodical manner.

This study provides statistically derived bounds for the uncertainty associated with surface temperature and precipitation measurements and yields a baseline dataset for validation of climate models as well as for a variety of other climatological uses. A two-step procedure using objective analysis was used to identify outliers. The first step was a temporal check that determines if a particular monthly value is consistent with other monthly values for the same station. The second step utilizes six different spatial interpolation techniques to estimate each monthly time series. Each of the methods is ranked according to its respective correlation coefficients with the actual time series, and the technique with the highest correlation coefficient is chosen as the best estimator. For both temperature and precipitation, a multiple regression scheme was found to be the best estimator for the majority of records. Results from the two steps are merged, and a combined set of quality control flags are generated.

Abstract

One of the major concerns with detecting global climate change is the quality of the data. Climate data are extremely sensitive to errant values and outliers. Prior to analysis of these time series, it is important to remove outliers in a methodical manner.

This study provides statistically derived bounds for the uncertainty associated with surface temperature and precipitation measurements and yields a baseline dataset for validation of climate models as well as for a variety of other climatological uses. A two-step procedure using objective analysis was used to identify outliers. The first step was a temporal check that determines if a particular monthly value is consistent with other monthly values for the same station. The second step utilizes six different spatial interpolation techniques to estimate each monthly time series. Each of the methods is ranked according to its respective correlation coefficients with the actual time series, and the technique with the highest correlation coefficient is chosen as the best estimator. For both temperature and precipitation, a multiple regression scheme was found to be the best estimator for the majority of records. Results from the two steps are merged, and a combined set of quality control flags are generated.

Save