This is the second review of ‘A Climate Data Record (CDR) for the global terrestrial water budget: 1984-2010’. I thank the authors for their response to reviewer questions and comments and for their revisions to the manuscript. A few minor points still remain to be clarified, however. These are detailed below.
Section 2.2.1, page 8-9: I thank the authors for their explanation in their response to the question about the use of “optimal” to describe the merging technique. Based on this, it seems that it is perhaps more accurate to qualify this “optimality” as constrained by or conditional on limited data availability rather than being the optimal approach. A few minor edits, consistent with the authors’ response to R1-Comment 2, in the abstract and section 2.2.1 would clarify this point.
Page 9, Lines 31-32: It would be helpful here to state that the assumed 10% error in VIC runoff is based on the authors’ experience and judgment “given there is no global grid level (0.5 degree in this study) runoff observations to quantify the error.”
Page 11, lines 25-30: I recognize the difficulty in accounting for groundwater extraction and management globally. As the authors note, there are regions (California Central Valley, US High Plains, Iran, etc) where historic regional storage declines challenge the assumption of zero long term TWSC. For completeness, a explanation of the consequences of this assumption on the CDR results for these areas, either in section 3.2 or in the discussion in section 5, seems warranted.
Page 13, line 22: In response to R1-comment 12, the authors state that the filtering out of basins with non-significant correlations “was done in order to remove those basins such as Indus and Senegal which might have incorrect observational data.” This was not immediately apparent from the explanation on page 13 of the revised manuscript – please edit to include this point for clarity.
Page 16, lines 8-12 and Figure S7: I thank the authors for the additional text and plots addressing the question of inter-annual variability in the CDR. However, this section is a little unclear. Please define SPI and provide more detail in the Figure S7 caption to indicate what is shown in the plots (i.e which parts are from the CDR?)
Figures – general: The number of figures in the main manuscript makes the main point of this manuscript less clear. Some of the figures could be moved to the supplemental information to better emphasize the results of the study. For example, the data product comparisons (Figures 2-10) could be limited to just continental (or river basin) plots and the remainder moved to the SI.
Figure 2: The caption references TMPART but this seems inconsistent. Should it instead reference the CSU dataset?
Figure 11: Please provide a more descriptive caption to accompany this flowchart. |