Dear reviewer,

Thanks for your valuable suggestions.

Yes, we will try to provide a more complete description of the EPGF tectonics and should also be able to provide other precipitation statistics.

yours

Hans-Balder Havenith , first author

Reply to Rev 1 comments :

"Authors report the landslides caused by Aug. 14, 2021 earthquake (Mw 7.2) in Haiti and

discuss the distributional and dimensional pattern with the landslides caused by Jan. 12,

2010 earthquake (Mw 7.2) in Haiti. Seismic, tectonic (briefly) and climatic (rainfall)

perspectives are also discussed to explain the role of these factors in landslide occurrence.

Though such types of base datasets are vital to understand the relative influence of

earthquakes and atmospheric-climatic phenomena (hurricanes), there are following

issues, which can further simplify the MS;

1) The epicentres in both the earthquake events; Aug. 14, 2021 earthquake (Mw 7.2) and

Jan. 12, 2010 earthquake (Mw 7.2) are related to Enriquillo-Plantain-Garden Fault

(EPGF), which is crucial in the genesis of these earthquakes. Though authors have

described the EPGF briefly in line no. 60-70, page no. 3, a thorough tectonic description

of the EPGF can be presented as supplementary annexure.

Answer by HBH (main author): A more detailed description of the  EPGF tectonics is being prepared and will be included in the revised version of the manuscript. 

2) In Fig. 7, authors have presented climatic, particularly rainfall spatial variability. Fig. 7bc

showing October rainfall variability follow the hypothesis relating Oct. 2016 hurricane

with the landslides generated in Oct. 2016 and then reactivated at larger scale in 14,

2021 earthquake (Mw 7.2). However, it would be better if the authors present rainfall

spatial variability of each month to justify the relative dominance of Oct. month

particularly in 2016.

Answer by HBH (main author): well, we possess yearly averaged data and we presented them - presenting

maps for all months would definitely exceed the number of pages allowed for the manuscript. 

However, we can include in the annex the August 2021 prec stats map marked by the Grace event - for comparison.

3) Authors have mentioned the landslide volume in absolute numbers in several

statements throughout the MS. Please explain how you determined the volume of these

landslides?

Answer by HBH (main author):  Thanks for this comment, indeed, we should have specified that we estimated

these volumes according to equations presented in Havenith et al. (2015) in Geomorphology. This information will be added

(plus example calculation, thickness estimate and related volume estimate).

Sincerely yours

HB Havenith, main author

Dear reviewer,

also these comments are all useful, but ...

Visits were mainly made along roads after the earthquake, most of which had been impacted by rock falls. 

Actually, zones affected by the numerous debris slides were barely accessible. (and are even less now due to the security situation in the country). We will include a related comment to explain the choice of the field photographs.

For the colour schemes ... we will re-re-recheck this issue (already highlighted by previous reviews), but as we have many maps, three different catalogues, this is really the best compromise we could find after dozens of tests. Maybe playing with image contrast could help improve the differentiation between the outlines and background colours. We'll try ..

yours

Hans-Balder Havenith, first author, on behalf of all co-authors.

Dear reviewer,

please, find below some more detailed answers to your comments (also to those included in the annotated manuscript):

The authors present a very detailed paper, based on numerous estimations derived from the modeling of recorded processes. In a multi-hazard framework, such evaluations are essential for developing proper disaster risk reduction measures, and the applied character of the paper enhances its importance, besides its fundamental one. 

The paper is generally well structured, but there are sometimes paragraphs which should better fit other chapters (especially related to the methodology and discussion-based ones, as outlined in the annotated .pdf). As well, sometimes the paper has paragraphs which are difficult to be followed, due to the length of the sentences and the permanent comparisons between events. 

1) Sometimes, the authors are using sentences looking like a report (e.g. pages 6-7).

Answers by HBH (main author): yes, we will revise some lengthy paragraphs and split them.

2) The graphics part is conclusive, but the choice of close colours in landslide delineation might not be the best one.

Answers by HBH (main author): as indicated in the quick answer part - we tried many combinations. We'll try again, but 

as we present multiple inventories, some compromise has to be made.

3) Also, field pictures (besides the RS ones) with debris flows (as recognized by the authors as being the most numerous landslides triggered by the 2021 earthquake; instead, only rock falls are shown) would enhance the reader's overall image. 

Answers by HBH (main author): Certainly you are right, but, here, we cannot provide any further answer than the one quickly given  - only roads in the affected are could be visited after the earthquake - and those were mostly affected by rockfalls. Debris slides occurred in ravines that were not accessible and cannot be visited anymore due to restricted access to the area for security reasons (closed due to riots).

See below answers to your main comments (not including simple corrections) in the annotated manuscript:

R1: Numerous paragraphs, especially in the second half of the intro, seem more suitable to "data&methodology" chapter.

HBH: We will check this issue and move parts to the next section.

R2: Streets reopening? Was this the reason?

HBH: We will specify that we started the landslide mapping task due to the evidence that many landslides had occurred in the target area.

R3: A rough size estimation would be helpful. 

HBH: we'll specify that landslides typically smaller then 500 m2 or ~2000 m3 could not be identified (see also comments in size-frequency part).

R4:  diregarding ... :  Is there a reason for this approach? If so, it should be stated.

HBH: we'll specified that these effects cannot be assessed as the frequency of aftershocks exceeds by far the one of available new remote imagery. Also,  most aftershocks had a M<4.5  .. which means that the likelihood for landslide triggering is small. This point will be more specified.

R5: What about the initiation of first time failures or reactivations of former landslide deposits, in areas not affected by the seismic shaking?

HBH: indeed, we did not map outside the 2010-2021 target zone, but at least over the area affected by the 2010 event and not by the 2021 one ..we could not detect any new landslides. We'll specify this point.

I think, above we answered all major comments of reviewer 2.

yours

Hans-Balder Havenith, main author.