Forecasting Rainfall IDF Curves Using Ground Data and Downscaled Climate Projections to Enhance Flood Management in Punjab, Pakistan

1. Summary

Thank you very much for taking the time to review this manuscript. Below, you will find our detailed responses to your comments, along with the corresponding revisions and corrections made in the manuscript. All modifications have been highlighted in green color within the text, to clearly indicate the changes made in response to the reviewers’ suggestions.

2. Point-by-point response to Comments and Suggestions for Authors

Comments 1: Page 4: Figure 1 has low resolution. As we zoom in on the image to identify

the cities, the resolution is lost. I suggest removing the names of the

cities that are not covered in the study, as well as highlighting only the

five cities selected for the study with outlines (thicker lines).

Response 1: We thank the reviewer for this valuable suggestion. Figure 1 has been updated with improved resolution to ensure better readability. Only the five selected study cities are now labeled, and their boundaries have been highlighted with thicker outlines for clear identification.

Comments 2: What is the meaning of the legend in the lower right corner of Figure 1? It

should be described in the text immediately below the figure.

Response 2: We appreciate the reviewer’s observation. The legend represents the Digital Elevation Model (DEM) in meters, illustrating elevation variability across Pakistan and Punjab. To enhance clarity, the legend label has been updated to “Elevation (m)” and a brief explanatory note has been added below the figure to describe the meaning of the DEM values, see lines 148-150.

Comments 3: Also regarding the study area, Pakistan is a populous country, so it is

important to present, either textually or in Figure 1, the population size of

each city. This information can be further explored in the conclusions.

Response 3: We thank the reviewer for this valuable suggestion. The population size of each selected city has been added directly to Figure 1 for better contextual understanding. Additionally, the importance of population concentration in relation to urban flood risk and infrastructure planning has been briefly discussed in the Conclusion section to highlight the practical implications of the findings (lines 698-710).

Comments 4: Line 292: “Equation (17) and (18) is known”. Correct to “Equations (17) and (18) are known”

Response 4: We appreciate the reviewer for noting this grammatical oversight. The sentence has been corrected to “Equations (17) and (18) are known” in the revised manuscript.

Comments 5: The conclusions can be improved with an additional paragraph addressing the possible consequences of the increased impact on the municipalities of Punjab. The Lahore region has a population of over 14 million, right? An increase in evolution and, consequently, the occurrence of natural disasters tend to be more catastrophic in this city, since the population density is higher among the studied locations.

These details provide a purpose to the work and help policymakers to have a

scientific basis for actions to mitigate the impacts of climate change.

Response 5: We sincerely thank the reviewer for this insightful suggestion. The Conclusion section has been revised to include implications of the projected rainfall intensification for the highly populated urban centers of Punjab. The new paragraph emphasizes the varying vulnerability levels among the studied cities and the importance of translating these findings into actionable strategies for flood resilience and climate adaptation, see lines 698-710.

Response to Reviewer 1 Comments

1. Summary

Thank you very much for taking the time to review this manuscript. Below, you will find our detailed responses to your comments, along with the corresponding revisions and corrections highlighted in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes

Are all the cited references relevant to the research?

Yes

Is the research design appropriate?

Can be improved

Suitable improvements have been incorporated.

Are the methods adequately described?

Can be improved

Suitable improvements have been incorporated.

Are the results clearly presented?

Can be improved

Suitable improvements have been incorporated.

Are the conclusions supported by the results?

Can be improved

Suitable improvements have been incorporated.

3. Point-by-point response to Comments and Suggestions for Authors

Comments 1: This study develops future projections of Intensity-Duration-Frequency (IDF) curves for major urban centers in Punjab, Pakistan using CMIP6 GCM simulation precipitation outputs. The title of this paper refers to Satellite-Based Climate  Projections, and this term is continuously used throughout the manuscript. The authors, based on published results, identified 8 GCMs which most adequately represent precipitation in Pakistan. These are NOT satellite-based data. The authors have misunderstood what CMIP6 models are. This aspect must be corrected. They also state that there is no best-model. Therefore, why do you select a single model to create IDFs, why not use the model ensemble considering the 8 models selected, and the model spread to evaluate uncertainty?

Response 1: We thank the reviewer for pointing out this concern. The statement “no single model excels universally” was intended to acknowledge that GCM performance varies regionally and that model suitability must be evaluated contextually. In this study, a comprehensive statistical ranking approach (using K–S, A–D, RMSE, and R²) was applied to all eight CMIP6 models to identify the one that best represents rainfall behavior in Punjab. The EC-Earth3-Veg-LR model consistently demonstrated superior performance across all stations and was therefore selected for IDF curve development. We agree that ensemble averaging can capture uncertainty but note that it also tends to smooth rainfall extremes, which are crucial for IDF analysis. Hence, a single high-performing model was adopted to preserve the realistic magnitude of short-duration rainfall events, while inter-model variability was assessed during the evaluation phase to ensure robustness.

Comments 2: Page 2 line 77 says “used GCMs and SSPs to assess future climate impacts”, why do you separate SSPs from GCMs, SSP is a boundary condition for the GCM. 

Response 2: We thank the reviewer for noting this technical clarification. The sentence has been revised to correctly state that Shared Socioeconomic Pathways (SSPs) serve as boundary conditions for the CMIP6 GCM simulations. The revised text now reads: “CMIP6 GCM outputs under different SSP scenarios to assess future climate impacts”. This correction ensures conceptual accuracy in describing the relationship between GCMs and SSP scenarios.

Comments 3: Page 2 Line 81 says “Similarly, Rohat et al. highlighted that rising rainfall extremes under SSP scenarios have major implications for stormwater system design [29]” believe it should be [19], which corresponds to Rohat et al.

Response 3: We thank the reviewer for catching this typographical error. The citation has been corrected, and Rohat et al. is now properly referenced as [19] in the revised manuscript.

Comments 4: Page 2 Line 86 says “GCMs provide satellite-based data for historical and future conditions…” what do you mean by satellite-based? CMIP6 GCMs are not based on satellite information

Response 4: We thank the reviewer for this clarification. The text has been revised to accurately describe that CMIP6 GCMs produce simulated precipitation data, which were subsequently bias-corrected using the satellite-derived observational dataset. The term “satellite-based” has been removed to avoid misunderstanding.

Comments 5: Figure 1. The names written on the map, I believe identifying states are unreadable. Please improve. Maybe number the states and create a list with the names. It is not clear if the map is of all Pakistan and Punjab is in green. If so, the legend might say

something like: Map of Pakistan, Punjab province in green. What do you mean by

“depicting the distribution…”, what distribution? The location of the 5 cities selected

should be clearly seen in this map

Response 5: We thank the reviewer for this valuable suggestion. Figure 1 has been updated to improve readability by increasing font size and clarity of district labels. The map now clearly highlights Punjab province within Pakistan, with the major study cities (Lahore, Sialkot, Multan, Bahawalpur, and Sargodha). The figure caption has also been revised for accuracy and clarity. The updated caption now reads: “Map of Punjab province (green color) in Pakistan, showing the district boundaries, elevation variation (DEM), and the locations of the five selected cities—Lahore, Sialkot, Multan, Bahawalpur, and Sargodha.”

Comments 6: Page 4. Line 156. Satellite-based is not correct. Are you using precipitation derived from satellite data? If not, omit using this term. CMIP6 are not based on satellite data

Response 6: We thank the reviewer for this observation. The term “satellite-based” has been removed throughout the manuscript to avoid confusion. As clarified in our response to Comment 1, CMIP6 GCMs provide simulated precipitation data that were bias-corrected using the CHIRPS satellite-derived observational dataset. Appropriate revisions have been made consistently across the text.

Comment 7: Line 162. GCMs do not forecast, this concept is incorrect. They only provide  information considering possible forcings, but it is not a forecast

Response 7: We appreciate the reviewer’s valuable clarification. The terminology has been modified throughout the manuscript to more accurately describe the nature of GCM outputs. As noted in our response to Comment 1, the study now consistently refers to future climate projections derived from CMIP6 GCM simulations under different SSP scenarios, rather than forecasts.

Comment 8: Line 165. Which satellite climate data were selected? No reference to specific satellite data is made in this paper. 

Response 8: We thank the reviewer for pointing this out. The omission of the specific dataset reference was unintentional. The satellite dataset used for this study—the Global Precipitation Climatology Project (GPCP) obtained from the Climate Data Store (CDS)—has now been clearly mentioned in the revised manuscript, as referred on lines 163-172.

Comment 9: Table 2. These are not satellite data, these are GCM models.

Response 9: We appreciate the reviewer for highlighting this inconsistency. The table title has been corrected from “Satellite datasets used in this research” to “General Circulation Models (GCMs) used in this research” to accurately reflect the nature of the listed datasets.

Comment 10: FIGURE 2. Why does the chart show 3 times the rectangle “check goodness of Fit and Develop empirical equation using Sherman method”? The previous steps use 1 rectangle but are applied to all 3 data groups, why the change? Same for “Develop functional relationships”

Response 10: We thank the reviewer for this insightful observation. The figure has been revised for clarity and consistency. The repeated rectangles were originally used to distinguish the separate processing paths for the observed, historical GCM, and future GCM datasets. However, we agree that “Check Goodness of Fit and Develop Empirical Equation using the Sherman Method” should be represented by a single unified step, as the same statistical evaluation process applies to all datasets.

In contrast, the Quantile Mapping step remains divided into two paths—spatial and temporal downscaling—to accurately reflect the methodological distinction between these two downscaling approaches. The updated figure now ensures both visual clarity and methodological accuracy, in line with standard EQM frameworks (as referenced in the supporting literature).

Comment 11: Line 210. Specify and describe the ranking method used. How are the values of the Final rank obtained? Why wasn’t R2  beside the final rank column used to define the final rank?

Response 11: We appreciate the reviewer’s insightful query. The ranking procedure has now been elaborated in the “Selection of Optimal GCM Model” subsection of the methodology. In summary, the performance of each GCM was evaluated for every city using three statistical goodness-of-fit tests (K–S, A–D, and RMSE). Each GCM was ranked according to its test statistic value for each metric, and the cumulative sum of these ranks determined the Final Rank—where the lowest total indicates superior model performance.

The R² values were calculated separately from scatter plots comparing observed and simulated rainfall to provide a supplementary measure of model agreement. However, R² was not included in the cumulative rank computation because the rank-based method relied solely on direct statistical fit indicators. This clarification and methodological explanation have been added to the manuscript.

Comment 12: Figure 5 does not add information, it could be omitted.

Response 12: We sincerely thank the reviewer for this valuable suggestion. We fully understand the concern regarding potential overlap with tabular data. However, we have retained Figure 5 as it provides a clear visual representation of the GCM performance rankings across all major cities of Punjab, allowing readers to quickly interpret inter-model variations and spatial consistency. We believe this visual summary complements the tabulated statistics by enhancing comprehension and supporting comparative analysis at a glance.

Comment 13: Table 5. Which are the units of the return periods?

Response 13: We appreciate the reviewer’s observation. The units of the return periods are in years, and this clarification has now been explicitly mentioned in the table caption and column heading to avoid any ambiguity.

Comment 14: Figure 6. Which are the units of the return periods?

Response 14: We thank the reviewer for this helpful comment. The units of the return periods are in years, and this has now been clearly indicated in the figure caption to ensure clarity for the readers.

Comment 15: Table 6. What is i? T? t? units?

Response 15: We appreciate the reviewer’s query. The table shows empirical formulas based on the Sherman equation, as explain in subheading 3.5 “Generating IDF Curves and Sherman Equation”. The variables have been explained while explaining the equation. In the formula “i” is rainfall intensity (mm/hr), “t” is storm duration (minutes), ”T” is return period (years). 

Comment 16: Figure 8. Shows historical GCM and Future GCM distributions, but future should be adjusted using the corrected GCM distribution, shown in Figure 7. The model is corrected in the present to be used to correct the future. The adjusted historical GCM in figure 7 increases extreme precipitation values, this is not considered when the future is corrected and there is a considerable reduction in extreme future precipitation

Response 16: We appreciate the reviewer’s insightful observation. The applied approach indeed follows the Equidistance Quantile Matching (EQM) method proposed by Srivastav et al. (2014), which first corrects the historical GCM data (spatial downscaling, Figure 7) and then applies this correction framework to the future GCM data (temporal downscaling, Figure 8).

While Figures 7 and 8 are presented separately for clarity—depicting the two stages of the EQM process—the temporal correction in Figure 8 inherently incorporates the bias adjustment derived from the corrected historical GCM distribution shown in Figure 7. This ensures that the improved historical bias structure is transferred to the future projections, maintaining consistency in the correction process.

4. Response to Comments on the Quality of English Language

Point 1: The English is fine and does not require any improvement.

Response 1: Thank you very much. We are glad to know that the quality of English in the paper is appropriate and as per requirements of the journal.

Thank you for the constructive feedback during the major revision process was invaluable, and we thank you again for your time and dedication in helping us improve our work for publication in the Atmosphere.

Response to Reviewer 2 Comments

1. Summary

Thank you very much for taking the time to review this manuscript. Below, you will find our detailed responses to your comments, along with the corresponding revisions and corrections highlighted in track changes in the re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes

Are all the cited references relevant to the research?

Yes

Is the research design appropriate?

Yes

Are the methods adequately described?

Yes

Are the results clearly presented?

Yes

Are the conclusions supported by the results?

Yes

3. Point-by-point response to Comments and Suggestions for Authors

Comments 1: Clarify the unique contribution of combining EQM downscaling with LP3-based probabilistic modeling. Is this the first attempt for Punjab or a methodological refinement of prior regional studies?

Response 1: Thank you for this important observation. We have added a focused clarification in the Results and Discussion section immediately after the LP3 best-fit declaration and the downscaling results [Page 21 - 22]. The added text explains that, while EQM and LP3 have been used separately in prior work, their integration — EQM to improve modelled inputs and LP3 to robustly model corrected extremes — provides a novel and practically useful pipeline for deriving station-specific IDF equations from CMIP6 projections in our study region. This clarification has been inserted to highlight the methodological contribution in the context of the actual results.

Comments 2: The rationale for selecting only five cities should be justified beyond historical flooding—e.g., data completeness or representativeness across climatic zones.

Response 2: Thank you for this thoughtful comment. The rationale for selecting the five study cities has been expanded in the revised manuscript (Data and Study Area section) [Page 4 - 5]. In addition to their flood susceptibility, these locations were chosen based on long-term data completeness, spatial coverage, and climatic diversity across Punjab. Each site has reliable rainfall records from PMD and represents a distinct hydrological setting within the province, ensuring that the analysis captures regional variability in rainfall extremes.

Comments 3: Results are extensive but largely descriptive. Include more quantitative comparisons (e.g., percentage differences between observed and downscaled datasets, performance statistics for EQM validation).

Response 3: We sincerely thank the reviewer for this valuable suggestion. The Results and Discussion sections have been substantially enhanced to include detailed quantitative comparisons. Specifically, percentage differences between observed, historical, and downscaled datasets have been incorporated, along with explicit RMSE and R² statistics for GCM evaluation and EQM validation. Quantitative improvements due to downscaling (ranging from 5–25% bias reduction) and projected increases in future rainfall intensities (20–55% under SSP scenarios) are now clearly presented and discussed. These additions strengthen the analytical depth of the manuscript and provide clearer evidence of the model’s performance and downscaling effectiveness.

Comments 4: The discussion should better connect findings to physical mechanisms — e.g., why short duration extremes increase disproportionately under SSP5-8.5

Response 4: We thank the reviewer for this insightful comment. The Discussion section [Page 29] has been refined to explicitly link the disproportionate increase in short-duration rainfall under SSP5-8.5 to enhanced convective activity, orographic influences, and the greater atmospheric moisture-holding capacity associated with a warmer climate, consistent with the Clausius–Clapeyron relationship (~7% per °C). These clarifications strengthen the physical interpretation of the results and align the discussion with established climate mechanisms driving rainfall intensification.

Comments 5: Uncertainty discussion is missing — particularly model spread, EQM limitations, and potential over-fitting.

Response 5: We thank the reviewer for highlighting this important aspect. A new paragraph has been added to the Discussion section [Page 29] acknowledging the key sources of uncertainty, including GCM model spread, EQM limitations related to bias stationarity, and the potential for overfitting due to data constraints. These uncertainties have been discussed along with a statement on the measures taken to minimize them (use of multiple GCMs, statistical validation, and cross-scenario consistency). This addition ensures a balanced interpretation of the results and strengthens the scientific rigor of the study.

4. Response to Comments on the Quality of English Language

Point 1: The English is fine and does not require any improvement.

Response 1: Thank you very much. We are glad to know that the quality of English in the paper is appropriate and as per requirements of the journal.

Thank you for the constructive feedback during the major revision process was invaluable, and we thank you again for your time and dedication in helping us improve our work for publication in the Atmosphere.

Response to Reviewer 2 Comments

1. Summary

Thank you very much once again for taking the time to review this manuscript in detail. Below, you will find our responses to your comments, along with the corresponding revisions and corrections highlighted in red in re-submitted files.

2. Questions for General Evaluation

Reviewer’s Evaluation

Response and Revisions

Does the introduction provide sufficient background and include all relevant references?

Yes

Is the research design appropriate?

Yes

Are the methods adequately described?

Can be improved

Suitable improvements have been incorporated.

Are the results clearly presented?

Can be improved

Suitable improvements have been incorporated.

Are the conclusions supported by the results?

Can be improved

Suitable improvements have been incorporated.

Are all figures and tables clear and well-presented?

Can be improved

Suitable improvements have been incorporated.

3. Point-by-point response to Comments and Suggestions for Authors

Comments 1: Satellite-based should be removed from the title, it is confusing

Response 1: We thank the reviewer for this valuable suggestion. The term “satellite-based” has been removed from the title to prevent confusion, as the study primarily relies on downscaled CMIP6 GCM projections rather than direct satellite observations. The revised title now reads:

“Forecasting Rainfall IDF Curves Using Ground Data and Downscaled Climate Projections to Enhance Flood Management in Punjab, Pakistan.”

This modification ensures greater technical accuracy and consistency with the data sources used in the study.

Comments 2: Section 3.6. You are talking about downscaling satellite data for future projections. I dont understand why you talk about satellite data, particularly because in the description of the methodology you talk about GCM and ground station data.

Response 2: We appreciate the reviewer’s careful observation. The section has been revised to clearly state that downscaling was applied to CMIP6 GCM precipitation data, not satellite data. The term “satellite data” has been replaced and rephrased for consistency with the methodology and data description. The updated section now emphasizes the use of corrected GCM outputs adjusted using observed ground station data through the EQM approach.

Comments 3: Lines 520-526. also talk about downscaling satellite data.

I believe the method is to downscale the models using satellite information.

Response 3: We thank the reviewer for the clarification. The text has been corrected to accurately describe that GCM simulated precipitation data were downscaled using the EQM method, with bias correction informed by observed and satellite-derived reference datasets. The revised lines now clearly specify that the downscaling was applied to GCM data, not to satellite data itself, while retaining all references and methodological details for clarity.

Comments 4: When you speak of observations, do you mean satellited-derived? Satellite precipitation is not observed, it is derived or synthetic precipitation. Satellites do not measure precipitation hence it cannot be considered an observation

Response 4: We thank the reviewer for this valuable clarification. The terminology has been refined throughout the manuscript to distinguish between ground-based observations and satellite-derived precipitation estimates. The term “observed data” is now exclusively used for rain gauge measurements, while references to satellite datasets have been revised to “satellite-derived” or “satellite-estimated” precipitation, ensuring technical accuracy and consistency with standard climatological definitions.

Comments 5: Figure 6 says observed data, are these local station precip data or satellite. Clarify in the legend- Same for Figure 7

Response 5: We thank the reviewer for pointing out this important clarification. The captions for Figures 6 and 7 have been updated to explicitly specify that the observed data refer to ground-based station measurements. The revised captions now read as follows:

·         Figure 6: IDF Curves using ground/observed data and historical GCM simulated data for (a) Bahawalpur, (b) Lahore, (c) Multan, (d) Sargodha, and (e) Sialkot (Return period in years).

·         Figure 7: Statistical relation between GCM simulated historical data and ground/observed data, to develop the spatially downscaled model output.”

Figures have been updated where possible to improve the readability and clarity for the reader throughout the manuscript. These changes ensure clear distinction between ground observations and model-simulated datasets.

Comments 6: Line 562. Same problem. Line 594. Same, the model is downscaled not the satellite information. Check for other parts in the paper with this problem

Response 6: We thank the reviewer for the observation. The terminology has been corrected throughout the manuscript to consistently specify that the downscaling has been updated accordingly to maintain clarity and consistency with the methodology description.

4. Response to Comments on the Quality of English Language

Point 1: The English is fine and does not require any improvement.

Response 1: Thank you very much. We are glad to know that the quality of English in the paper is appropriate and as per requirements of the journal.

5. Additional clarifications

-