You are here

Due to System Maintenance there may be a delay in order processing.

RATPAC-A

Radiosonde Atmospheric Temperature Products for Assessing Climate dataset A (RATPAC-A) contains adjusted global, hemispheric, tropical, and extratropical mean temperature anomalies. ​A collaborative effort by NOAA scientists collected observations at 85 stations using hydrogen-filled weather balloons equipped with a radiosonde, providing data from 13 atmospheric pressure levels and at near-global coverage. RATPAC-A is one of two distinct datasets that emerged from a need for a radiosonde time series that was less influenced by inhomogeneities. Scientists used a procedure called first differencing to reduce the inhomogeneities in this dataset. 

From 1958 through 1995, the bases of RATPAC-A are on spatial averages of the Lanzante et al. (2003; hereafter LKS) adjusted 87-station temperature data. After 1995, they are based on the Integrated Global Radiosonde Archive (IGRA) station data, combined using a first difference method (Free et al. 2004). For analyses of interannual and longer-term changes in global, hemispheric, and tropical means, the NOAA team recommends using RATPAC-A because it contains the most robust large-scale averages. Access to the RATPAC-A data files is available on the FTP site.

First Difference Method

For RATPAC-A, the LKS data used the first difference (FD) method. Discussed in more detail in the article "Using First Differences to Reduce Inhomogeneity in Radiosonde Temperature Datasets," the design of this method is to reduce inhomogeneities in large-scale mean time series without adjusting the individual time series. The method involves taking the difference in temperature between one time step and the next (the "first difference"), then computing large-scale means of the FD series, and finally reconstructing large-scale temperature series from the FD series. Omitting portions of the station time series around the times of known changes in instruments or procedures, is an attempt to eliminate the effect of inhomogeneities due to such changes. However, the method introduces a random error that increases with the number of time gaps in the data and increases with decreasing number of stations, so the results are limited to large-scale means. Although this method does not use neighbor stations as reference series in the usual sense, it does rely on other stations in a region to supply information about temperature change at times of metadata events at an affected station. This method does not adjust individual stations.

The first difference method was applied to the IGRA monthly means starting in 1996. Before 1996, the RATPAC-A time series is the mean of the data modifications for both liberal and conservative (LIBCON) adjusted LKS station data, without versions of the LKS adjusted data, and without the use of FD. Note that the LKS authors preferred using the “LIBCON” subset; one of several available versions of the LDS adjusted data.

Although the LKS dataset runs through 1997, substituting IGRA data for 1996 and 1997 appeared to be the best solution because the short record left after the adjustments makes LKS adjustments in 1996 and 1997 less reliable than previous years. The LKS approach requires several years of data both before and after a possible inhomogeneity to make the best adjustment.

Starting in 1995, the team removed six months of data from the IGRA monthly means before and after each metadata event for any station having a relevant event documented by a report from the country in which the station was located. Some events were considered relevant only for certain stations and levels. For example, a change in reporting practices affecting temperatures below -90°C was considered relevant for stations and levels where temperatures near -90°C were reported. Data removal occurred at times in 1996 and 1997 where LKS had made adjustments or had removed data due to homogeneity concerns. Despite recent efforts by NCEI to update the station histories, useful metadata after 1995 was available for just 38 of the 85 stations. Based on this metadata, NCEI removed 29 stations for that time frame. Combination of the series occurred using the method that is described in more detail in the Appendix of "Radiosonde Atmospheric Temperature Products for Assessing Climate (RATPAC): A new data set of large-area anomaly time series."

Spatial Averaging

In an effort to obtain spatially unbiased large-scale means, the team compensates for uneven longitudinal distribution of stations by creating regional means before averaging data into zonal bands. Each 30-degree zonal band was divided into three longitudinal regions of 120 degrees each: 30°W to 90°E, 90°E to 150°W, and 150°W to 30°W. Hemispheric (0°–90°), tropical (30°S–30°N), and extratropical (30°–90°) means were calculated from these zonal means, alreally weighted using the cosine of the latitude of the midpoint of the zone and the global mean was the average of the hemispheric means. To facilitate comparison with other datasets, NCEI provides the time series for the region from 20°N to 20°S.

Endpoint Outlier Trimming

The FD procedure introduced an endpoint outlier trimming procedure to reduce the random errors. As described in Peterson et al. (1998) and Free et al. (2004), this procedure removes data exceeding a prescribed multiple of the standard deviation of the original time series if the data fall at the end of a data segment (immediately before or after a gap). Using a larger multiple as a cutoff rather than a smaller multiple, the removal of fewer data points occurs. Results from the FD procedure are sensitive to the choice of this multiple, or trim factor (see Free et al. 2005). For the reasons outlined in that paper, the team chose a factor equal to one standard deviation from the mean.

Interpolation

In another effort to reduce random errors, the LKS authors used linear interpolation between months of data to fill data gaps of less than four months before removing data at the times of metadata events. The use of interpolation is only at stations where data removal occurred due to metadata events. The mean number of months of data added by interpolation to these 38 stations was about six per station.

References

Free M., D.J. Seidel, J.K. Angel, J. Lanzante, I. Durre and T.C. Peterson, 2005: Radiosonde Atmospheric Temperature Products for Assessing Climate (RATPAC): A new dataset of large-area anomaly time series. Journal of Geophysical Research, 110, D22101, http://dx.doi.org/10.1029/2005JD006169.

Free, M., J.K. Angel, I. Durre, J. Lanzante, T.C. Peterson and D.J. Seidel, 2004: Using first differences to reduce inhomogeneity in radiosonde temperature datasets. Journal of Climate, 17, 4171-4179, http://dx.doi.org/10.1175/JCLI3198.1.

Lanzante, J.R., S.A. Klein, and D.J. Seidel, 2003: Temporal homogenization of monthly radiosonde temperature data. Part I: Methodology. Journal of Climate, 16, 224-240, http://dx.doi.org/10.1175/1520-0442(2003)016<0224:THOMRT>2.0.CO;2.

Peterson, T.C., T.R. Karl, P.F. Jamason, R. Knight, and D.R. Easterling, 1998: First difference method: Maximizing station density for the calculation of long-term global temperature change. Journal of Geophysical Research Atmospheres, 103, 25967-25974, http://dx.doi.org/0.1029/98JD01168.