Journal cover Journal topic
Earth System Dynamics An interactive open-access journal of the European Geosciences Union

Journal metrics

  • IF value: 3.635 IF 3.635
  • IF 5-year<br/> value: 3.869 IF 5-year
    3.869
  • CiteScore<br/> value: 4.15 CiteScore
    4.15
  • SNIP value: 0.995 SNIP 0.995
  • SJR value: 2.742 SJR 2.742
  • IPP value: 3.679 IPP 3.679
  • h5-index value: 21 h5-index 21
https://doi.org/10.5194/esd-2017-81
© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 4.0 License.
Research article
25 Sep 2017
Review status
This discussion paper is a preprint. It is a manuscript under review for the journal Earth System Dynamics (ESD).
Bias correction of surface downwelling longwave and shortwave radiation for the EWEMBI dataset
Stefan Lange Potsdam Institute for Climate Impact Research, Telegraphenberg A 31, 14473 Potsdam, Germany
Abstract. Many meteorological forcing datasets include bias-corrected surface downwelling longwave and shortwave radiation (rlds and rsds). Methods used for such bias corrections range from multi-year monthly mean value scaling to quantile mapping at the daily time scale. An additional downscaling is necessary if the data to be corrected have a higher spatial resolution than the observational data used to determine the biases. This was the case when EartH2Observe (E2OBS; Calton et al., 2016) rlds and rsds were bias-corrected using more coarsely resolved Surface Radiation Budget (SRB; Stackhouse Jr. et al., 2011) data for the production of the meteorological forcing dataset EWEMBI (Lange, 2016). This article systematically compares various parametric quantile mapping methods designed specifically for this purpose. The methods vary in the time scale at which they operate, in their way of accounting for physical upper radiation limits, and in their approach to bridging the spatial resolution gap between E2OBS and SRB. It is shown how temporal and spatial variability deflation related to bilinear interpolation and other deterministic downscaling approaches can be overcome by downscaling the target statistics of quantile mapping from the SRB to the E2OBS grid such that sub-SRB-grid scale spatial variability present in the original E2OBS data is retained. Cross-validations at the daily and monthly time scale reveal that it is worthwhile to take empirical estimates of physical upper limits into account when adjusting either radiation component and that, overall, bias correction at the daily time scale is more effective than bias correction at the monthly time scale if sampling errors are taken into account. A validation against independent ground observations from the Baseline Surface Radiation Network (BSRN; König-Langlo et al., 2013) suggests that the bias correction of E2OBS surface downwelling radiation using SRB data that was done for the production of EWEMBI had a positive and neutral overall effect on rlds and rsds, respectively. Using any of the other methods tested here would have given similar results as the biases relative to BSRN remaining after bias correction are dominated by the corresponding SRB data biases.

Citation: Lange, S.: Bias correction of surface downwelling longwave and shortwave radiation for the EWEMBI dataset, Earth Syst. Dynam. Discuss., https://doi.org/10.5194/esd-2017-81, in review, 2017.
Stefan Lange
Stefan Lange

Data sets

EartH2Observe, WFDEI and ERA-Interim data Merged and Bias-corrected for ISIMIP (EWEMBI)
S. Lange
https://doi.org/10.5880/pik.2016.004
Stefan Lange

Viewed

Total article views: 217 (including HTML, PDF, and XML)

HTML PDF XML Total BibTeX EndNote
187 27 3 217 3 2

Views and downloads (calculated since 25 Sep 2017)

Cumulative views and downloads (calculated since 25 Sep 2017)

Viewed (geographical distribution)

Total article views: 217 (including HTML, PDF, and XML)

Thereof 213 with geography defined and 4 with unknown origin.

Country # Views %
  • 1

Saved

Discussed

Latest update: 17 Oct 2017
Publications Copernicus
Download
Short summary
The bias correction of surface downwelling longwave and shortwave radiation using parametric quantile mapping methods is shown to be more effective (i) at the daily than at the monthly time scale, (ii) if the spatial resolution gap between the reference data and the data to be corrected is bridged in a more suitable manner than by bilinear interpolation, and (iii) if physical upper limits are taken into account during the adjustment of either radiation component.
The bias correction of surface downwelling longwave and shortwave radiation using parametric...
Share