| 23 Sep 2025
Increased heating of the land surface as hot-dry events persist
Abstract. Compound hot-dry events have devastating effects on ecosystems as well as societies. Combinations of more incoming shortwave radiation (SW_down) and drying soil moisture lead to the build-up of high temperatures during dry periods. In this process, evaporation (ET) plays an important role in coupling temperature and soil moisture, and thus can lead to feedback loops and more drying. While both atmospheric contributors (SW_down) and the land surface (ET) are known to influence temperature during dry periods, it remains unclear how their relevance for high temperatures varies throughout a dry event, i. e. from the build-up of heat to its persistence during ongoing dryness. Furthermore, the contributions of ET and SW_down to heat onset and persistence during dryness are likely to differ across space and over the last decades. In this study, we investigate SW_down and ET changes as two contributing factors to heat accumulation throughout dry events and across recent decades using reanalysis data. We determine periods of soil dryness accompanied by high temperatures using weekly timescale data. Within the detected hot-dry weeks, we distinguish between heat onset and heat persistence by evaluating the continuity of high temperatures. By mapping changes in ET and SW_down during heat onset and heat persistence, we find that radiation increases contribute to the onset of heat globally but are less dominant for heat persistence. Evaporative cooling mitigates radiation-driven temperature increases during the onset of heat in humid regions. By contrast, this effect vanishes during persistent high temperatures. While the general occurrence of hot conditions during dry events increased from 14 % to 28 % from the 1980s to the 2010s, the evolution of ET and SW_down throughout hot-dry events shows no clear trend over the last few decades. Our study emphasizes that contributors to heat development and/or persistence vary during the lifetime of a dry event which should be considered when analyzing compound extremes.
Summary:
This study examines how atmospheric forcing and land-surface processes interact to shape the development and persistence of high temperatures during dry events, based on reanalysis data from the 1980s to the 2010s. By distinguishing between heat onset and heat persistence within soil moisture defined dry periods, the authors assess the respective roles of incoming shortwave radiation and evaporation. They find that the fraction of dry events accompanied by high temperatures has increased markedly, although no clear long-term trend emerges in the evolution of shortwave radiation and evaporation during these events. The analysis suggests that increased radiation primarily contributes to the onset of heat, while evaporative cooling can mitigate early heat development in humid regions but loses influence during persistent or more arid hot-dry conditions.
Overall comments: The manuscript addresses a timely and relevant question regarding the drivers of heat onset and persistence during dry events; however, the current methodological framework is not sufficiently robust to support the strength of the attribution claims made. The analysis relies largely on descriptive week-to-week changes, which limits the ability to draw firm conclusions about relative importance or causality. As presented, the reasoning linking changes in shortwave radiation and evaporation to temperature evolution remains somewhat speculative and not adequately supported by rigorous statistical testing or sensitivity analyses. The exclusive focus on two drivers is also insufficiently justified, particularly given the well-established roles of atmospheric circulation, advection, cloud dynamics, and other land-atmosphere feedback in shaping heat extremes. Several spatial patterns are interpreted qualitatively without adequate quantitative substantiation. Furthermore, key methodological definitions require clarification, and the presentation of figures and language would benefit from substantial revision. In its current form, the study does not provide sufficiently strong analytical evidence to substantiate its central conclusions, and major methodological and conceptual revisions is necessary.
Major Comments:
The methodology appears insufficient for the attribution claims made in the paper. The identification of heat onset and persistence relies primarily on percentile-based thresholds, but the attribution of drivers is based largely on descriptive analysis. The conclusions regarding the relative importance of different drivers would be more convincing if supported by more robust statistical or causal analyses.
The authors rely mainly on shortwave radiation and evapotranspiration as drivers of heat onset and persistence. However, several other variables can influence these processes, such as vapor pressure deficit (VPD), soil temperature, ground evaporation, atmospheric circulation indices etc. The authors should justify why only two variables were selected or expand the analysis to include additional relevant drivers.
The manuscript acknowledges that processes such as advection and atmospheric circulation may drive heat persistence. However, these processes are not explicitly analyzed. The authors should include at least some atmospheric circulation or advection metrics to demonstrate their potential role.
A more formal causal attribution could be performed using: Partial correlation analysis, Multiple regression, or other statistical attribution approaches. This would allow the authors to quantify the relative importance of different drivers, especially in the identified hotspots. The role of vegetation is only represented through tree cover fraction. However, vegetation structure and biome type are likely to influence the response of heat and drought events. The analysis could be extended to comparison of responses across biomes (e.g., forests, grasslands, croplands) and provide spatial maps showing structural or biome-dependent effects. This would add ecological context and strengthen the interpretation of the results.
Several statements describing spatial patterns are qualitative and would benefit from quantitative support. For example: Line 135: Identification of regional hotspots, Line 149: Relationship between net radiation and shortwave radiation, Line 157: Regional differences in ET, SWdown, and soil moisture responses. These statements should be accompanied by quantitative metrics, regional statistics, comparative tables or plots
The manuscript attributes some of the radiation changes to cloud dynamics, for example: Line 166: Dissipation of clouds leading to increased SWdown. Line 239: Minor cloud cover reducing SWdown during persistence. However, these statements are speculative and not supported by explicit analysis. The authors should include a cloud-related variable (e.g., cloud fraction or radiation components) to demonstrate this mechanism quantitatively.
The relationship between heat onset and drought onset is not clearly explained. The authors should clarify: Whether heat onset corresponds to the beginning of drought. Or whether heat onset is defined independently of drought development. This distinction is important for interpreting compound events.
In Figure 2, the authors indicate that the median was used for all analyses (Figure 3-4). It would be helpful for the authors to clarify why the median was selected instead of other statistical measures. A brief explanation, particularly regarding the data distribution, presence of outliers, or other methodological considerations, would improve clarity and strengthen the justification of the approach.
No time series are shown to illustrate the temporal evolution of key variables across decades or within individual dry events. Including time series plots of temperature, shortwave radiation, and evapotranspiration would improve transparency and help substantiate the claims about temporal dynamics and trends.
The histograms currently present only the global distribution. Including separate histograms for different regions would enable clearer comparison of regional patterns. These additional figures could be provided in the appendix to maintain the flow of the main text while offering more detailed regional insights.
The definition of aridity (Table 2) should be clearly explained in the Methods or Results section before the term is introduced. The exact range used for classification also needs to be specified to ensure clarity and reproducibility. Additionally, including a map that illustrates the spatial distribution of aridity would facilitate clearer comparison across regions and enhance the overall presentation of the results.
Minor Comments:
The overall language in the manuscript appears somewhat casual, and there are a few grammatical and punctuation issues. The authors are encouraged to revise the manuscript carefully for language accuracy. Punctuation issues are noticeable in Lines 123, 125, 171, and 197.
Some sentences require rephrasing for clarity and readability. For example:
Line 51: “For example, atmospheric circulation creates persistent weather patterns leading to the build-up of high temperatures (Brunner et al., 2017; Barriopedro et al., 2023). On a regional scale, advection of warm air can further increase temperatures during dry periods (Miralles et al., 2019; Schumacher et al., 2019) and secondly, diel dynamics in the local boundary layer support the persistence of high temperatures by trapping and reintroducing warm air to the surface-near air (Miralles et al., 2014).”. The description of atmospheric circulation, advection, and boundary-layer processes is long and difficult to follow. The sentence should be broken into shorter, clearer statements.
Line 193: “Independent of what causes the temperature increases resulting from increased H during dryness - i) increases in SW_down, ii) SW_down increases in combination with SM-suppressed ET or iii) other factors - magnitudes are comparable” The sentence describing different mechanisms leading to temperature increases during dryness is complex and difficult to interpret. It should be simplified and restructured for clarity.
The figure captions require further development, as they are currently too brief and do not provide sufficient detail for readers to fully understand the figures. Detailed captions should clearly describe the content, variables, and key elements of each figure which would improve clarity and enhance the overall readability of the manuscript.
Figure 1: Use a consistent font style throughout the figure. The sensible heat flux label is currently overlapping with the clipart, which reduces clarity. Also, clearly define what “T” represents in the figure. Consider making the labels bold, increasing the line thickness, and enlarging and bolding the arrows to improve readability. These adjustments will make the figure clearer and more visually effective.