Identification of compound drought and heatwave events on a daily scale and across four seasons

Authors: Baoying Shan et al.

Egusphere-2023-147

This paper discusses ways to determine heatwaves, droughts, and combination of them in a systematic way and at daily scale allowing to better capture the start and end date of the events. They use 120 years of data over Uccle (point) and suggested an objective method to remove minor spells and merge mutually dependent ones and discussed four ways to define a CDHW event. They eventually used the outcomes to study the seasonality of the drought, heatwaves, and combination of them. They found that seasonal patterns, with the occurrences of droughts and heatwaves being negatively dependent in winter but positively dependent in the other three seasons at Uccle.

I believe the paper is well structured and well written in almost all aspects: the science within the scope of the journal, relatively novel in the method and concepts, important conclusion, good literature review, and good discussion of the outcomes and associated interpretation. The conclusions are backed up with proper studies, the title and abstract are fine.

I only have few minor points:

• The study is performed using a point measurement that has 120 years of daily temperature and precipitation data. Yet, the number of events, especially compound events, are not much. Is there any recommendation for the minimum length of the data record if someone wants to apply the method over other regions?
• Line 26: remove “more” before regionally
• Line 39: what is (7d)? Fig 7d in their paper?
• Line 98: not sure how you come up with 91 days? Is the 91 day window for both 30yr and entire record analysis? I think it is better to further clarify this.
• Line 224: shouldn’t it be the sum of “positive” SHI?
• Line 254: Check if the orders is right. I think it should be 1.3, 1, and 0.5 instead.
• Explain why no merging is carried out in some cases in Table 1.
• For Nd=1 and SPI =-1.3, what does Md=51 mean? why all other Mds are much less? Are these results stable?

This study presents a method to identify compound drought/heatwave daily events as well as develop a set of indices that can be used to characterize such events. The paper is well organized and the topic is appropriate for the journal. However, there are a couple of issues that should be addressed before publication (along with some other minor issues outlined below).

In particular, given the definition of drought and heatwave events I'm not sure that a "precise identification of the start and end dates of an event" is really realistic, I would probably focus more on things like duration etc. which are important and somewhat remove the aspect of a specific start/end date. Also, there is not much discussion on validating the identification of these compound events and although it is a difficult task I would have liked to see some additional text/analysis. A discussion section is also missing that would add some more details on how the results compare in the context of other studies; also, the implications of identifying the compound events can be expanded upon slightly (as the authors seem to attempt in the final sentence of the Conclusions section).

- l. 29: is the "coarse resolution" referring to spatial, temporal or both? From the next sentence it seems it's temporal but it'd be good to be explicit here.

- Introduction is otherwise well written.

- l. 76-80: I don't find this adds much value. Instead, an overview of the experimental design and its rationale would work better here.

- l. 82: why was only one meteorological station used in the study?

- l. 84: perhaps use "have been used in many studies" instead of "have been subjected to many studies"?

- l. 132-134: to what degree does the choice of distribution change the calculation of the indices? The use of AIC is an objective way of selecting the distribution, but what is the sensitivity of SPI/SHI?

- l. 148: the etymology of the term "deficiency" refers to quality, I'm wondering if there could be a better term to describe "how close two neighboring spells are" (unless I misunderstood the definition). Perhaps "proximity"?

- l. 168: I would remove the word "Concretely" here, perhaps "Particularly"?

- l. 230: not sure you need a separate subsection for this.

- l. 309: why specifically these values for Nd, Nh?

- l. 360-374: the Conclusions section is somewhat poorly written. For instance, the last sentence seems out of place while the statement "The results confirmed the effectiveness of the proposed method" needs a bit more context.