Assessing Vegetation Response to Multi-Scalar Drought across the Mojave, Sonoran, Chihuahuan Deserts and Apache Highlands in the Southwest United States

Round 1

Reviewer 1 Report

I have no further comments, the authors have fully addressed my comments on the previous version.

Reviewer 2 Report

I recommend this manuscript to be accepted in present form. The topic of this study is very interesting, since climate change is something that everyone faces across the globe. The better understanding of climate change can be done through research studies that take into account historical data from long periods, like this one. Finally, the combined use of indices such as SPEI and NDVI can provide useful insights for finding solutions related with the mitigation of climate change impact and must be further investigated.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Khatri-Chhetri et al. investigate the relationship between NDVI and an indicator of drought over the SW United States. They investigate the impacts of different time scales of drought/precipitation on vegetation growth. The paper has potential, but the authors need to clarify their methods and language. For example, how were the SPEI data integrated over the different time periods (1,2 month, etc)? Was it an average? The authors need to use consistent terminology and improve the clarity of text in many sections.

As a minor aside, most sources list the “Chihuahuan” not “Chihuahua” Desert.

Line 24: What sensor did the NDVI come from? Important to know what scale your data represents in the abstract.

Line 38: I think the term “Southwest” is too generic as a keyword. Would “Southwest US” be better?

Lines 70-73: The purpose of this sentence is unclear. “in different ecoregions…with different time lags”, can the authors be more specific? Is the point that the three scales (site, ecoregion, and biome) respond differently to drought?

Line 79: Different timescales? Do the authors mean different patterns of water availability as in the previous sentence?

Line 89: “four ecoregions in the Southwest U.S. to shed new…”

Line 93: I think it would be helpful for the authors to more clearly define what they mean by “timescales.” My assumption is that the authors would be looking at the impacts of 6-month duration drought, 1-year, 5-year, etc. Or is it the lagged response of vegetation to drought? More explicitly state the intent here or at the first occurrence of this term.

Line 99: Add a spatial qualifier “in deserts of the Southwest US”

Line 111: List the source of your ecoregion data? EPA? What level is the ecoregion data?

Line 120: With higher summer (winter) precip in the east and (west).

124: Legume trees: be more specific. Acacia and mesquite.

Lines 128-134: Was the analysis focused on two months based on data limitations or for ease of interpretation? Certainly there some locations (e.g. high elevations) that wouldn’t have peak greenness in either of the analyzed months.

Figure 1: 1) would be nice to provide an inset map showing the location of the study area within the U.S. 2) What is the source of the vegetation cover data provided in the map? Is it source 22 or 23?

Figure 2: Are the data presented in this figure based on an average across multiple years? If so, define which years.

Lines 147-148: Why did the authors chose to define winter as January to June and summer as July to December? Most sources list water year precipitation as October-September. Would it not follow then, that winter precipitation would be Oct-Mar, and summer as Apr-Sept?

Line 156: Why did you exclude data from Aqua? For a longer time-series? What collection of MODIS data did you use?

Line 174: Wet or dry relative to that site right? In other words looking across time?

Line 186: What not longer timescales as well? 2-year, 3-year, etc.?

Line 200: Wouldn’t it have been better to start with a higher resolution dataset (e.g. 30-m NLCD), then resample to 250m?

Line 228-229: So the correlation would be the monthly average? NDVI from April (or Sept) vs the average? SPEI from the previous 1 month, 2 months, 3 months, etc? This needs to be made more clear.

Line 239: How do the authors determine what the “dominant” timescale is? What do they mean by dominant?

Line 259: What does this negative trend actually indicate? Plotting all the 1,2,6,12 month mean SPIE values vs time?

Lines 260-261: Since this result was found in all ecoregions, replace “in these ecoregions” with “across the study area.”

Line 263: become

Line 267: To be consistent in terminology, replace “precedent” with “historic”

Lines 269-270: If more intense and longer dry periods are occurring in the study period, how is there no variation in stationarity of the SPEI data as described in the previous paragraph?

Figure 5: Why is the missing data for some time periods (i.e. ~2012 in the Chihuahuan)? Also, are these yearly averages plotted? It seems that the temporal density is higher than this (i.e. variation can be seen at a sub-annual timescale).

Lines 293-295: I thought (from lines 192-198) that the historical data was also SPEI, not SPI. Why is this procedure “we used the drought severity classification of Standardized Precipitation Index (SPI) to quantify SPEI used by U.S. Drought Monitor1” not indicated in the methods section?

How was SPEI calculated from SPI?

Lines 299-300: Reword to: “The Mojave Desert has the highest percentage of significantly correlated pixels among all ecoregions during the winter (Figure 8: b).”

Line 302: What is the “dominant” SPEI timescale? Not clear what is meant here, be more explicit. See previous comment.

Line 310: Reword to “The number? of pixels in which NDVI and SPEI were strongly correlated was…”

Sec 3.2.1 and 3.2.2 Though these sections are labeled as “April” and “September” it wouldn’t hurt to make this clear in text.

Lines 319-320: The sentence “the significantly correlated….” Is awkward and needs to be reworded., same for lines 321-322 and 323-324. Suggest rewording as on my comment on Line 310.

Figure 7: The caption isn’t clear. Suggest rewording to: “The spatial distribution of pixels with significant (p ≤ 0.05) correlations between vegetation productivity (NDVI) and varying SPEI timescales (in months) for April….”

Figure 7:  Figure 1 shows an “other” land cover class, I assume this is urban, water, agriculture, etc. Are these areas included in analysis? If not, they need to be clearly indicated in Figure 7 to differentiate them. If so, this decision needs to be justified.

Also “highest negative” is a bit odd, replace with “strongest negative”

Line 346: Delete “of”

Lines 345-351: Shouldn’t this text be pointing to Figure 8, not 9?

Line 357-358: This sentence is awkward, reword.

Lines 361 and 352: Delete “the” before “grassland” and before “sparse”

Line 378: Delete “different”

Line 391-393: This isn’t really a result. Delete or rephrase.

Line 394: Delete “in these regions”

Line 396-397: This sentence is too obviously true following the preceding sentence, delete.

Line 397: SPEI wasn’t “associated” with the higher water stress, the authors have been using these terms interchangeably throughout the paper. With this being the case, SPEI was the higher water stress.

Lines 398-400: How is this sentence: “In line…” not redundant with lines 393-394?

Lines 400-401: Again, this sentence “These dry…” seems redundant with lines 397-398: “The decreasing trend….” Combine or delete.

Line 405: Or shifts in the “boundaries” among ecoregions. Or more clearly, the shifts in species ranges.

Lines 407-408: This is at least the third time you’ve defined NDVI and water stress. No need to do that so many times. Would be preferable to just stick with just NDVI and SPEI, and avoid usage of other terms to define the same.

Also, this sentence unnecessarily repeats information that was presented earlier in the paper.

Line 412: “The seasonal precipitation….precipitation gradients…” is awkward. Rephrase and avoid using the same term twice in a sentence.

Line 425-426: Again, it has already been well established what the winter and summer seasons are.

Line 432: events tend not event tends.

Line 433: delete “that helps to increase their productivity” it distracts from the point of the sentence.

Lines 437-440: a) There should be at least 3 sentences in a paragraph, b) why not just make a new section heading and delete these sentences?

Line 441-443: If this is case, then why the need to analyze the differences between ecoregions?

Lines 448-450: This sentence “In contrast…” is a repeat of information presented in lines 431-434.

Line 455: Be clear, “slower” compared to what?

Line 461: Starting a paragraph with “however” feels quite odd in this case. “However” in response or rebuttal to what?

Lines 467-469: No need for the “however” I don’t think these are conflicting results at all. Kemp simply showed the vegetation is most productive in early spring and later summer, you are showing to which time period of precipitation this growth is correlated. Not a contradiction of previous results from what I can see.

Line 477-478: drier and warmer = summer, wet and cold =winter. Better to use consistent terminology throughout the paper, this improves reader comprehension.

Lines 509-510: How could this information reduce drought impacts? I don’t see how it could. Rather, it could help managers understand or manage drought impacts.

Line 515: Its not just drought impacts. It’s the impact of when precipitation occurs, and the lagged response of vegetation to precipitation.

Author Response

Thank you for your feedback, comments, and suggestions on our manuscript! Please see the attachment to find our reply to your comments.

Thank you again!

Author Response File: Author Response.pdf

Reviewer 2 Report

The only connection of this paper to the Remote Sensing journal is about its use of MODIS NDVI. Otherwise, it reads more fit within other climate change or eco-hydrology journals. So, I would recommend trying the current version on journals other than the RS. When you try a different journal, I would suggest the following minor changes:

1. have the study period extended to 2000-present day;
2. put figures 1-3 into SM, and;
3. only select 1-2 figures for the section 3.1;  

Author Response

Thank you for your feedback, comments, and suggestions on our manuscript! Please see the attachment to find our reply to your comments.

Thank you again!

Author Response File: Author Response.pdf