In the paper “Comparisons of satellites liquid water estimates to ECMWF and GMAO analyses, 20th century IPCC AR4 climate simulations, and GCM simulations” by Li et al. (Geophysical Research Letters, 35, L19710, doi:10.1029/2008GL035427, 2008, hereinafter LET), LET provided an assessment of the representation of the cloud liquid water path (LWP) by general circulation models (GCMs), namely the GCMs that contributed to the Coupled Model Intercomparison Project 3 (CMIP3) and that were utilized in the IPCC 4th Assessment. Since the publication of LET, discrepancies have been revealed in the interpretation of the request for model output and model documentation. Here we report a revision to the analysis results of LET based on correcting for these discrepancies. The corrected results illustrate smaller root mean square errors than that presented in LET and significantly improve the results of two GISS models and the IPSL model. While the results presented for a few specific models change, the main conclusions of LET remain valid, namely that significant disagreements of LWP are found among the models including both widely varying magnitudes and considerable differences in spatial patterns.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 The study by Li et al. [2008, hereinafter LET] examined the fidelity of 16 CMIP3 [Meehl et al., 2007; Intergovernmental Panel on Climate Change, 2007] models to represent cloud liquid water path (LWP) for 20th century simulations. The overall findings of the study were that there is considerable disagreement amongst the modeled values of LWP estimates. Typically, GCM LWPs have been found to be much larger than the observed estimates from CloudSat [Li et al., 2008]. Since the publication of this study, we have learned that there were inconsistencies in the interpretation of the relevant model output quantity, namely “clwvi”. LET's interpretation of this quantity was that it represented the vertical integral of cloud liquid water (LWP) and thus was the complement to the “clivi” variable specified to be the vertical integral of the cloud ice water (IWP). It turns out that the intended meaning of “clwvi” is total water path, i.e., ice plus liquid. Our follow-up investigation has determined that thirteen of the GCMs (labeled inLET as: ccc_ma63, cnrm, csiro, gfdl, iap, ipsl, mirochr, gisseh, gisser, inmcm, mpi, ukmogem and ukmocm) provided output that was consistent with the intended interpretation latter (i.e., the total water path (TWP)), three of the GCMs (labeled in LETas: bccr, csiro, ncar ) provided the output with the interpretation that the quantity was just associated with liquid water (i.e., LWP). Two GCMs output from ukmocm3 and ukmogem, the TWP to CMIP3 did not include water associated with the convection scheme. The purpose of this article is to provide a correction of the original analyses. For this purpose, we only replot the globally-averaged, long-term annual mean bar chart and the long-term annual mean maps; the original versions fromLET are plotted in Figures 1a and 1b, respectively. Here we show LWP directly for those models that output it as such or derive it for the models that provided TWP and IWP (i.e., LWP = TWP − IWP).
2. Results and Summary
Figure 2 shows a bar chart (Figure 2a) of the globally-averaged, long-term annual mean LWP and horizontal maps (Figure 2b) of long-term annual mean LWP from the GCMs analyzed inLET. The GCM data are based on the 1970–94 period of the 20th century GCM simulations (20c3m scenario). It is evident that model-to-model disagreements for globally-averaged cloud LWPs are similar to those in Figure 1a ofLETas well as the disparity in spatial patterns of the time-mean values shown in Figure 1b ofLET. The biggest correction is associated with the two GISS GCMs and IPSL GCM which when accounting for removing the IWP exhibits relatively small values, rather than large as before, compared to the other CGCMs. The RMS difference of the globally-averaged, long-term mean LWP fromLET is 51 (g m−2), while in the corrected set presented in Figure 2 is 44 (g m−2). To summarize the results of this note, the two GISS models and IPSL model results show much better results and the bias/RMS of CMIP3 LWP reduced. The impact of the correction, however, has little impact on the qualitative findings of LET, namely that there are still substantial differences (e.g., order of magnitude) in both LWP magnitudes and spatial patterns amongst the models. The significant disagreement among models for such a fundamental quantity, which has important ramifications in the context of climate change, must be reduced to improve future model projections of climate.