Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Summary 

This manuscript makes a valuable contribution to understanding hydroclimatic trends in Texas watersheds. It investigates long-term hydroclimatic trends in the Navasota River Watershed (NRW) in east-central Texas, specifically examining before and after the Lake Limestone dam (constructed in 1978). The authors use both station data and PRISM gridded precipitation data (validated against observations) with streamflow records . The methodological framework employs autocorrelation analysis, Mann-Kendall and Modified Mann-Kendall trend tests, and Pettitt’s change-point detection.

Key findings include a significant increasing trend in decadal precipitation averages, rising 10-year moving averages of peak streamflow. While no statistically significant change points were detected, it highlights a persistent, gradual intensification of flood risks, potentially driven by land-use changes and seasonal precipitation shifts (specifically in June).

 

 A General concept comments focusing scientific content

Statistical Significance & Interpretation (Crucial):

• Pettitt’s Test Misinterpretation (Section 3.6 & Discussion): The manuscript reports p-values of 0.83 and 0.796 for the Pettitt’s test (Figure 8). These values indicate NO statistically significant change points. However, the discussion (e.g., Line 535) treats 2011 and 2021 as meaningful "potential turning moments." This misrepresents the statistical results. If it is not significant, it is inappropriate to suggest these years represent a regime shift. Please revise the text to clearly state that no regime shift was detected.
• Edge Effect Artifacts: The detection of a change point at the very end of the data period (2021) in a dataset ending in 2021 is statistically unreliable and likely an artifact (edge effect). Please double-check the validity of this result or explicitly acknowledge this limitation in the text.
• Trend Line Description (Line 418): For Figure 7b, the text correctly notes that Kendall’s Tau (0.079) and the p-value (0.250) indicate an insignificant trend. However, the subsequent sentence describes a "gently rising trend." This is contradictory and potentially misleading. It is better to strictly describe this as "non-significant" rather than implying a visual trend exists where there is no statistical evidence.

Attribution of Hydroclimatic Changes:

• Teleconnections vs. Local Factors (Lines 470–472, 515–517): The discussion focuses heavily on the Dam and land cover changes. However, given the multi-decadal scope of the study, have you considered the impact of broader tele-climate oscillations (e.g., Pacific Decadal Oscillation (PDO) or ENSO)? Discussing these natural climate cycles is essential to distinguish them from anthropogenic local impacts.
• Lake Limestone Dam Details: Since the Dam is a keyword and a central focus of the "flood hazard" hypothesis, the introduction requires more specific detail. Please include the dam's storage capacity and a brief history of its operations. How specifically does the dam influence downstream flood hazards? Are streamflow dependent to the water release from the dam?

PRISM Data Validation:

• Independence Check: When validating PRISM data against station observations (NMD and CSTN), please confirm whether these specific stations were used by the PRISM team to generate the gridded dataset. If they were included in the model, the validation is circular.

Land Cover Methodology:

• Section 2.3: Please describe the specific methodology used to analyze the Land Cover data. The paper jumps to results (Figure 10) without explaining how the NLCD data was processed or compared.

 

Specific comments referring to line numbers, tables, or figures

Line 41: Spell out CONUS for international readers.

Line 55: The phrase "xxx km² below Lake Limestone" is confusing. Does this refer to a specific sub-basin area? Or should it be a distance?

Line 67: The numbering of the third hypothesis should be (3), not (2).

Line 84: Remove "(NRW)" as the abbreviation was already defined in Line 54.

Lines 90–91: Suggest changing to "...using the modified Mann-Kendall trend test."

Line 100: There is a discrepancy in watershed area. Line 55 mentions 4,073 km², but here it shows 5,822 km². Please ensure consistency or clarify the difference between the sub-basin and total watershed.

Line 106: Please use a comma for the thousand’s separator: 1,118 mm.

Lines 117–119: Suggest rephrasing for clarity: "...the rainfall stations at Navarro Mills Dam (NMD) and College Station Eastwood Field (CSTN) (green circles), as well as the streamflow gauging station of the Navasota River near Easterly (red circle)."

Lines 126–127: Please spell out PRISM at the first mention. It would be better to move the detailed description of PRISM to after Line 133 (after the flow data introduction) to improve flow.

Line 135: Spell out USGS for international readers.

Line 179: "Spatial differences" might be misleading here. Since you are comparing different time periods, should this header be "Spatiotemporal precipitation changes"?

Lines 212–214: Keep terminology consistent for readability. The first sentence refers to "flows," while the second switches to "wet years/dry years." Using "high flows/low flows" throughout would be smoother.

Lines 219–221: You mention the Modified Mann-Kendall test here but haven't defined it yet. Please describe the difference between the conventional and modified tests before this point.

Lines 252–253: Does your use of the Modified Mann-Kendall test refer to the method by Hamed and Rao (1998)? Please cite.

Section 2.3.2 and 3.2: Clarify in the text that you are applying the change analysis to each rainfall pixel

Figure 1:

• Where is the Lake Limestone Dam located on the map?
• The caption uses "Watershed," but the legend uses "Basin." Please be consistent.
• The station CSTN appears to be missing from the map.

Figure 3:

• Please use the same color scale across all four maps to allow for direct comparison.

Figure 6:

• The labels (a)–(d) are unclear/hard to read.
• For panel (c), please add (mm) to the axis label.

Figure 7 & 8:

• Clarify Legends: The legends are confusing. Is (a) yearly average streamflow and (b) yearly peak streamflow? Please specify this clearly in both the caption and the legend.
• Units: "Cumecs" is informal. Please use the standard scientific unit.
• Figure 7(c) & 8(b): Clarify if these are "moving averaged peak streamflow."

 

Author Response

REVIEWER 1

 RESPONSE TO THE REVIEW of the manuscript:

 Revealing Emerging Hydroclimatic Shifts: Advanced Trend Analysis of Rainfall and Streamflow in the Navasota River Watershed

Manuscript ID: hydrology-4007227

Dear Reviewer

Thank you for your time and effort in reviewing the manuscript. The comments were extremely helpful in improving the manuscript. Below we have repeated the comments and immediately following it, both our response and the improved text in the revised manuscript. The revised text is in blue.

 

Thanks again for your time and effort.

General comment

This manuscript makes a valuable contribution to understanding hydroclimatic trends in Texas watersheds. It investigates long-term hydroclimatic trends in the Navasota River Watershed (NRW) in east-central Texas, specifically examining before and after the Lake Limestone dam (constructed in 1978). The authors use both station data and PRISM gridded precipitation data (validated against observations) with streamflow records. The methodological framework employs autocorrelation analysis, Mann-Kendall and Modified Mann-Kendall trend tests, and Pettitt’s change-point detection.

Key findings include a significant increasing trend in decadal precipitation averages, rising 10-year moving averages of peak streamflow. While no statistically significant change points were detected, it highlights a persistent, gradual intensification of flood risks, potentially driven by land-use changes and seasonal precipitation shifts (specifically in June).

Response: Thank you for your general comment. This manuscript makes a valuable contribution to understanding hydroclimatic trends in Texas watersheds. Your comment helps validate the relevance of our findings for flood risk assessment and watershed management in Texas, particularly in the context of potential land-use changes and shifting seasonal rainfall patterns. We have ensured that the revised manuscript clearly articulates these implications and strengthens the narrative around the gradual but meaningful hydroclimatic shifts observed in the NRW.

Major Comments

Statistical Significance & Interpretation (Crucial):

Comment 1: Pettitt’s Test Misinterpretation (Section 3.6 & Discussion): The manuscript reports p-values of 0.83 and 0.796 for the Pettitt’s test (Figure 8). These values indicate NO statistically significant change points. However, the discussion (e.g., Line 535) treats 2011 and 2021 as meaningful "potential turning moments." This misrepresents the statistical results. If it is not significant, it is inappropriate to suggest these years represent a regime shift. Please revise the text to clearly state that no regime shift was detected.

Response: Thank you for this thoughtful observation. While addressing Comment 2, we identified an issue in our initial Python implementation of the Pettitt test and corrected it before re-running the analysis. The updated results indicate that the annual peak streamflow series does not contain a statistically significant change point (p ≈ 0.30), suggesting that years such as 2011 and 2021 should not be interpreted as regime shifts. In contrast, the 10-year moving average of annual peak streamflow reveals a highly significant change point around 1990 (p ≈ 1.63 × 10⁻⁶). This multi-decadal shift aligns with the precipitation patterns described in Section 3.2 and occurs roughly twelve years after the construction of Lake Limestone Dam, suggesting a reservoir-related influence on long-term hydrologic behavior. Sections 3.6 and 4.4 have been revised to reflect these corrected findings and to ensure that the discussion accurately represents the statistical results. Kindly see our revision for sections 3.6 and 4.4 following comment 2.

Comment 2: Edge Effect Artifacts: The detection of a change point at the very end of the data period (2021) in a dataset ending in 2021 is statistically unreliable and likely an artifact (edge effect). Please double-check the validity of this result or explicitly acknowledge this limitation in the text.

Response: Thank you for this important observation. In revisiting our Pettitt test results, after correcting our implementation as described in the response to Comment 1, we confirmed that the previously reported change point at the end of the record (2021) was indeed an edge-effect artifact rather than a meaningful hydrologic shift. As you noted, detecting a change point at the final year of a dataset, especially when using a moving-average series, is statistically unreliable.

To address this, we re-ran the Pettitt test on both the raw annual peak streamflow series and the 10-year moving-average series after removing edge-affected years. In both cases, the end-of-record change point disappeared and was not statistically significant. The only robust change point identified was the highly significant shift around 1990 (p ≈ 1.63 × 10⁻⁶) in the moving-average series, which aligns with the long-term hydrologic patterns described in Sections 3.2 and 4.1. Kindly see our revision for sections 3.6 and 4.4 below.

Comment 3: Trend Line Description (Line 418): For Figure 7b, the text correctly notes that Kendall’s Tau (0.079) and the p-value (0.250) indicate an insignificant trend. However, the subsequent sentence describes a "gently rising trend." This is contradictory and potentially misleading. It is better to strictly describe this as "non-significant" rather than implying a visual trend exists where there is no statistical evidence.

Response: Thank you for this helpful observation. We agree that describing the trend as “gently rising” is misleading given the non-significant Kendall’s Tau and p-value. The paragraph has been revised to clearly state that the trend is statistically non-significant and that the plotted line should not be interpreted as evidence of a meaningful increase. Kindly see the revised paragraph below.

Attribution of Hydroclimatic Changes:

Comment 4: Teleconnections vs. Local Factors (Lines 470–472, 515–517): The discussion focuses heavily on the Dam and land cover changes. However, given the multi-decadal scope of the study, have you considered the impact of broader tele-climate oscillations (e.g., Pacific Decadal Oscillation (PDO) or ENSO)? Discussing these natural climate cycles is essential to distinguish them from anthropogenic local impacts.

Response: Thank you for this insightful comment. We agree that considering large-scale climate oscillations is essential given the multi-decadal scope of the study. The paragraphs have been revised to explicitly acknowledge the role of teleconnections such as ENSO and PDO and to clarify how these broader climate drivers interact with local factors, such as the dam and land cover change, in shaping the observed hydroclimatic patterns. Kindly see our revision below.

Comment 5: Lake Limestone Dam Details: Since the Dam is a keyword and a central focus of the "flood hazard" hypothesis, the introduction requires more specific detail. Please include the dam's storage capacity and a brief history of its operations. How specifically does the dam influence downstream flood hazards? Are streamflow dependent to the water release from the dam?

Response: Thank you for this valuable comment. We have revised the introduction to include the dam’s storage capacity, its operational history, and a clearer explanation of how reservoir releases may influence downstream hydrologic conditions. These additions strengthen the context for understanding the dam’s role in the study’s flood hazard hypothesis. Kindly see our revision below.

PRISM Data Validation

Comment 6: Independence Check: When validating PRISM data against station observations [Navarro Mills Dam (NMD) and College Station Eastwood Field (CSTN)], please confirm whether these specific stations were used by the PRISM team to generate the gridded dataset. If they were included in the model, the validation is circular.

Response: Thank you for raising this important point. We have verified that both Navarro Mills Dam (NMD) and College Station Eastwood Field (CSTN) are included among the stations assimilated by the PRISM system in generating the gridded precipitation dataset. Therefore, using these same stations for validation would introduce circularity. In response, we have revised the sentence to indicate that the comparison was used to assess consistency. Kindly see the revised sentence below.

Land Cover Methodology

Comment 7: Section 2.3: Please describe the specific methodology used to analyze the Land Cover data. The paper jumps to results (Figure 10) without explaining how the NLCD data was processed or compared.

Response: Thank you for your important comment. In this study, we did not perform additional land-cover classification or apply new processing techniques beyond the standard products provided by the National Land Cover Database (NLCD). Instead, we downloaded the NLCD datasets for the relevant years and used them directly to evaluate land cover change dynamics within the Navasota River Watershed. The classification methodology, accuracy assessments, and processing procedures are those already conducted by the NLCD team. To address your concern, we have clarified this point in the revised manuscript and added cross-references to Section 2.2.3, where the NLCD classification system, accuracy information, and the relevant documentation are described. This ensures transparency regarding our use of the dataset and prevents confusion about whether additional classification or processing steps were undertaken. Kindly check section 2.2.3 in the revised manuscript.

Specific comments referring to line numbers, tables, or figures

Comment 8: Line 41: Spell out CONUS for international readers.

Response: Thank you for this suggestion. We have revised the manuscript to spell out Continental United States (CONUS) upon first use to ensure clarity for international readers.

Comment 9: Line 55: The phrase "xxx km² below Lake Limestone" is confusing. Does this refer to a specific sub-basin area? Or should it be a distance?

Response: Thank you for pointing out this ambiguity. We have clarified that 4,073 km² refers to the total watershed area located downstream of Lake Limestone, not a distance measurement. The revised text now clarifies this point more clearly to avoid confusion. Please see our revision following the response to comment 5.

Comment 10: Line 67: The numbering of the third hypothesis should be (3), not (2).

Response: Thank you for noting this error. We have corrected the numbering so that the third hypothesis is now properly labeled as (3) in the revised manuscript.

Comment 11: Line 84: Remove "(NRW)" as the abbreviation was already defined in Line 54.

Response: Thank you for your observation. We have removed the repeated abbreviation “(NRW)”, since it was already defined earlier in the manuscript.  

Comment 12: Lines 90–91: Suggest changing to "...using the modified Mann-Kendall trend test."

Response: Thank you for the suggestion. We have revised the text to read “…using the standard and modified Mann–Kendall trend test.” Kindly see our revision below.

Comment 13: Line 100: There is a discrepancy in watershed area. Line 55 mentions 4,073 km², but here it shows 5,822 km². Please ensure consistency or clarify the difference between the sub-basin and total watershed.

Response: Thank you for bringing this discrepancy to our attention. We have clarified that 5,822 km² corresponds to the total watershed area, whereas 4,073 km² refers only to the downstream portion below Lake Limestone. The sentences have been revised accordingly for consistency. Kindly see our revision below.

Comment 14: Line 106: Please use a comma for the thousand’s separator: 1,118 mm.

Response: Thank you for the correction. We have updated the value to 1,118 mm using the proper thousand’s separator.

Comment 15: Lines 117–119: Suggest rephrasing for clarity: "...the rainfall stations at Navarro Mills Dam (NMD) and College Station Eastwood Field (CSTN) (green circles), as well as the streamflow gauging station of the Navasota River near Easterly (red circle)."

Response: Thank you for the suggestion. We have revised the sentence for clarity as recommended. Kindly see our revision below.

Comment 16: Lines 126–127: Please spell out PRISM at the first mention. It would be better to move the detailed description of PRISM to after Line 133 (after the flow data introduction) to improve flow.

Response: Thank you for the helpful suggestion. We have spelled out PRISM at its first occurrence for clarity. Regarding the placement of the PRISM dataset description, we appreciate the recommendation; however, we believe the current location maintains a clearer logical flow within our methodology section. Therefore, we have retained the description in its original position while ensuring the text is streamlined for readability.

Comment 17: Line 135: Spell out USGS for international readers.

Response: Thank you for the suggestion. We have spelled out United States Geological Survey (USGS) upon first mention to ensure clarity for international readers.

Comment 18: Line 179: "Spatial differences" might be misleading here. Since you are comparing different time periods, should this header be "Spatiotemporal precipitation changes"?

Response: Thank you for the clarification. We agree that “Spatial differences” may be misleading in this context. We have revised the header to “Spatiotemporal Precipitation Changes” to more accurately reflect the comparison of different time periods. Kindly see the revised section title below.

Comment 19: Lines 212–214: Keep terminology consistent for readability. The first sentence refers to "flows," while the second switches to "wet years/dry years." Using "high flows/low flows" throughout would be smoother.

Response: Thank you for this helpful suggestion. We have revised the terminology to maintain consistency by using high flows/low flows throughout the description. Kindly see our revision below.

Comment 20: Lines 219–221: You mention the Modified Mann-Kendall test here but haven't defined it yet. Please describe the difference between the conventional and modified tests before this point.

Response: Thank you for this helpful suggestion. Instead of adding a full description at this point of the text, we now direct readers to Section 2.3.4, where the differences between the standard and Modified Mann–Kendall test are clearly explained. This approach avoids redundancy and maintains a logical flow in the manuscript. Kindly see the directing sentence below.

Comment 21: Lines 252–253: Does your use of the Modified Mann-Kendall test refer to the method by Hamed and Rao (1998)? Please cite.

Response: Thank you for bringing this to our attention. Yes, our use of the Modified Mann–Kendall test refers to the method developed by Hamed and Rao (1998). We have added the appropriate citation in the revised manuscript to ensure proper attribution.

Comment 22: Section 2.3.2 and 3.2: Clarify in the text that you are applying the change analysis to each rainfall pixel

Response: Thank you for this helpful suggestion. We have revised Sections 2.3.2 and 3.2 to explicitly clarify that the precipitation change analysis was applied to each PRISM rainfall pixel. This ensures transparency in the methodology and helps readers understand how the spatial maps were generated. Please see our revision below.

Comment 23: Figure 1: Where is the Lake Limestone Dam located on the map? The caption uses "Watershed," but the legend uses "Basin." Please be consistent. The station CSTN appears to be missing from the map.

Response: Thank you for this important observation. We have revised Figure 1 to address the indicated issues. Kindly see the updated figure below.

Comment 24: Figure 3: Please use the same color scale across all four maps to allow for direct comparison.

Response: Thank you for this valuable comment regarding the colormap interpretation in Figure 3. We agree that, in some visualizations, similar hues can create ambiguity between positive and negative changes. To minimize this risk, Figure 3 uses a stretched color scheme with decade-specific value ranges, and each panel includes its own explicit legend showing the exact minimum and maximum percentage changes for that decade. Because the magnitude and distribution of rainfall change vary substantially from the 1980s to the 2010s, using separate legends ensures that the colormap accurately reflects the spatial variability within each decade rather than forcing all panels into a single uniform range that could suppress meaningful detail.

Comment 25: Figure 6: The labels (a)–(d) are unclear/hard to read. For panel (c), please add (mm) to the axis label.

Response: Thank you for bringing this to our attention. We have improved the visibility and clarity of the labels (a)–(d) in Figure 6, and we have added “(mm)” to the axis label in panel (c) as requested. Kindly see the updated figure below.

Comment 26: Figure 7 & 8: Clarify Legends: The legends are confusing. Is (a) yearly average streamflow and (b) yearly peak streamflow? Please specify this clearly in both the caption and the legend. Units: "Cumecs" is informal. Please use the standard scientific unit. Figure 7(c) & 8(b): Clarify if these are "moving averaged peak streamflow."

Response: Thank you for these helpful suggestions. According to your suggestions, we have revised Figures 7 and 8 to improve clarity and consistency. Kindly see the updated figures below.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper investigated trends and changing points of rainfall and streamflow data in the Navasota River Watershed (NRW) in Texas. The authors found positive trends in precipitation and streamflow over the basin, which was then attributed to dam construction and land cover changes. The paper is comprehensive in its trend analysis using established methods, and its findings are meaningful for local water management. The objectives, methods, and results are clearly described and presented, along with detailed discussion and interpretation. However, the analysis mainly attributes the positive trends to dam and land cover shifts, which neglects the impacts of broader climate change (e.g., temperature increases) or teleconnections (e.g., ENSO). I suggest major revision before this paper can be accepted, as a more thorough discussion of these major climate drivers is needed for understanding extreme events in this watershed.

1. Line 333 and Figure 4c: Comparing the average precipitation from 1900 to 1978 to only the year of 2015 seems unfair to support the trends of precipitation. Extreme precipitation of a single year can be influenced by climate variability. In this case, the year of 2015 is an extreme El Nino event, where a persistent atmospheric blocking pattern was formed over the U.S. west coast that led to exceptionally high rainfall over Texas (Chen & Kumar, 2018). It is unfair to attribute the rainfall changes only to dam construction or land cover changes, as the ENSO played a primary role. The impact of climate change and ENSO should be included and discussed.
2. Line 346: The autocorrelation coefficient is 0.04 for lag 1, which is inside the shaded area and seems to be statistically insignificant. This contrasts to the statement in the paragraph that the lag-1 autocorrelation is significant. Please clarify this. If there is no significant autocorrelation, the statement about temporal memory of the watershed (e.g., line 569-570) should be removed.
3. Section 3.4: Would be interesting to see if there is a trend in the timing of extreme precipitation or discharge, and whether the trend is significant.
4. Line 541: “The watershed might be going through a gradual change in hydrologic regime”: However, the results in section 3.2 show different rainfall patterns over pre- and post-dam periods (line 327-328). Comparing rainfall before and after dam construction, together with the discussion in section 4.1, implies that the construction of the reservoir can pose a shift in local rainfall and hydrological patterns. This seems to be contradictory to the discussion in section 4.4, where it suggests that there are no significant change points and the trends are gradual.

 

Minor comments:

1. Figure 3: The colormap seems misleading as the yellow region can indicate both positive and negative changes. The first instinct is that the yellow colors indicate decreasing rainfall while green and blue indicate increasing rainfall.
2. Figure 5: Provide definition of the shaded area in the figure caption.
3. Line 202: When mentioning streamflow data, please clarify whether it is annual maximum discharge or annual mean discharge.

Reference:

Chen, M., & Kumar, A. (2018). Winter 2015/16 Atmospheric and Precipitation Anomalies over North America: El Niño Response and the Role of Noise. Monthly Weather Review, 146(3), 909–927. https://doi.org/10.1175/MWR-D-17-0116.1

Author Response

REVIEWER 2

 RESPONSE TO THE REVIEW of the manuscript:

Revealing Emerging Hydroclimatic Shifts: Advanced Trend Analysis of Rainfall and Streamflow in the Navasota River Watershed

Manuscript ID: hydrology-4007227

Dear Reviewer

Thank you for your time and effort in reviewing the manuscript. The comments were extremely helpful in improving the manuscript. Below we have repeated the comments and immediately following it, both our response and the improved text in the revised manuscript. The revised text is in blue.

 

Thanks again for your time and effort.

General comment

The paper investigated trends and changing points of rainfall and streamflow data in the Navasota River Watershed (NRW) in Texas. The authors found positive trends in precipitation and streamflow over the basin, which was then attributed to dam construction and land cover changes. The paper is comprehensive in its trend analysis using established methods, and its findings are meaningful for local water management. The objectives, methods, and results are clearly described and presented, along with detailed discussion and interpretation. However, the analysis mainly attributes the positive trends to dam and land cover shifts, which neglects the impacts of broader climate change (e.g., temperature increases) or teleconnections (e.g., ENSO). I suggest major revision before this paper can be accepted, as a more thorough discussion of these major climate drivers is needed for understanding extreme events in this watershed.

Response: Thank you for your general comment. We agree that the original discussion placed too much emphasis on dam operations and land cover change. In the revised manuscript, we have expanded the Discussion to include the roles of broader climate drivers, such as regional warming and major teleconnections (e.g., ENSO, PDO), and clarified how these interact with local factors to influence precipitation and streamflow trends in the NRW.

 

Major comments

Comment 1: Line 333 and Figure 4c: Comparing the average precipitation from 1900 to 1978 to only the year of 2015 seems unfair to support the trends of precipitation. Extreme precipitation of a single year can be influenced by climate variability. In this case, the year of 2015 is an extreme El Nino event, where a persistent atmospheric blocking pattern was formed over the U.S. west coast that led to exceptionally high rainfall over Texas (Chen & Kumar, 2018). It is unfair to attribute the rainfall changes only to dam construction or land cover changes, as the ENSO played a primary role. The impact of climate change and ENSO should be included and discussed.

Response: Thank you for this important observation. We agree that comparing the long-term pre-dam average (1900–1978) to a single year like 2015 is not an appropriate basis for identifying long-term precipitation changes. As you noted, 2015 was a strong El Niño year, during which large-scale atmospheric circulation anomalies were the primary drivers of extreme rainfall in Texas (e.g., Chen & Kumar, 2018). In the revised manuscript, we emphasized the influence of ENSO and climate variability when discussing 2015 rainfall and expanded the discussion to incorporate the broader roles of climate change and teleconnections (ENSO, PDO) in shaping precipitation trends in the NRW. Kindly see our revision in the results and discussion section reflected in the paragraphs below.

Comment 2: Line 346: The autocorrelation coefficient is 0.04 for lag 1, which is inside the shaded area and seems to be statistically insignificant. This contrasts to the statement in the paragraph that the lag-1 autocorrelation is significant. Please clarify this. If there is no significant autocorrelation, the statement about temporal memory of the watershed (e.g., line 569-570) should be removed.

Response: Thank you for bringing this discrepancy to our attention. You are correct that the lag-1 autocorrelation coefficient (r = 0.04) falls within the confidence bounds and is statistically insignificant. In the revised manuscript, we have corrected the text to state that no significant autocorrelation was detected at lag 1, and we have revised statements implying temporal memory or persistence in the watershed. This revision ensures that our interpretation aligns accurately with the autocorrelation results. Kindly see our revision for section 3.3 below.

Comment 3: Section 3.4: Would be interesting to see if there is a trend in the timing of extreme precipitation or discharge, and whether the trend is significant.

Response: Thank you for this thoughtful suggestion. In the current study, we focused on the magnitude-based trends in precipitation and streamflow; however, we did not specifically analyze trends in the timing of extreme precipitation or peak discharge events. We agree that examining shifts in event timing (e.g., earlier or later seasonal occurrence of extremes) could provide valuable insight into changing hydrologic behavior. While such an analysis is beyond the scope of the present study, we have added a statement in the Conclusion noting this as an important direction for future research and recommending that future studies evaluate whether the timing of extremes exhibits significant long-term trends. Kindly see the added sentence below.

Comment 4: Line 541: “The watershed might be going through a gradual change in hydrologic regime”: However, the results in section 3.2 show different rainfall patterns over pre- and post-dam periods (line 327-328). Comparing rainfall before and after dam construction, together with the discussion in section 4.1, implies that the construction of the reservoir can pose a shift in local rainfall and hydrological patterns. This seems to be contradictory to the discussion in section 4.4, where it suggests that there are no significant change points and the trends are gradual.

 Response: Thank you for this thoughtful observation. We appreciate the reviewer’s attention to consistency between the rainfall analysis in Section 3.2, the hydrologic context discussed in Section 4.1, and our original interpretation in Section 4.4. In response to a similar comment from reviewer 1, we conducted a full review of our Pettitt test implementation and re-evaluated the streamflow datasets using revised Python code.

Our updated analysis yielded results that differ from the earlier version:

• While annual peak streamflow values do not exhibit a statistically significant change point (p ≈ 0.30),
• The 10-year moving average of annual peak streamflow reveals a highly significant change point in 1990 (p ≈ 1.63 × 10⁻⁶), which is visible in the revised Figure 8b.

This hydrologic shift aligns closely with the patterns described in Section 3.2, where we documented differences in rainfall behavior between the pre-dam (before 1978) and post-dam periods, as well as with the streamflow patterns discussed in Section 4.1. The identified change point in 1990 occurs approximately twelve years after the construction of Lake Limestone Dam, suggesting that the reservoir has played a meaningful role in altering the downstream hydrologic regime.

Accordingly, we have revised Sections 3.6 and 4.4 to reflect this updated interpretation. The revised text clarifies that, although annual peak flows remain highly variable with no abrupt statistical break, the longer-term moving-average behavior demonstrates a significant, reservoir-influenced shift in hydrologic conditions beginning around 1990. This harmonizes the discussion across Sections 3.2, 4.1, and 4.4 and provides a coherent explanation linking rainfall patterns, dam construction, and observed hydrologic changes.

Please see the updated versions of Sections 3.6 and 4.4 below.

Minor comments

Comment 5: Figure 3: The colormap seems misleading as the yellow region can indicate both positive and negative changes. The first instinct is that the yellow colors indicate decreasing rainfall, while green and blue indicate increasing rainfall.

Response: Thank you for this valuable comment regarding the colormap interpretation in Figure 3. We agree that, in some visualizations, similar hues can create ambiguity between positive and negative changes. To minimize this risk, Figure 3 uses a stretched color scheme with decade-specific value ranges, and each panel includes its own explicit legend showing the exact minimum and maximum percentage changes for that decade. Because the magnitude and distribution of rainfall change vary substantially from the 1980s to the 2010s, using separate legends ensures that the colormap accurately reflects the spatial variability within each decade rather than forcing all panels into a single uniform range that could suppress meaningful detail.

Comment 6: Figure 5: Provide definition of the shaded area in the figure caption.

Response: Thank you for this suggestion. We have updated the caption for Figure 5 to clarify the definition of the shaded region. Kindly see the updated Figure 5 caption below.

Comment 7: Line 202: When mentioning streamflow data, please clarify whether it is annual maximum discharge or annual mean discharge.

Reference: Thank you for the comment. We have clarified in the revised manuscript that the streamflow dataset refers to annual averages, peaks, and 10-year moving averages. Kindly see our revision of the sentence below.  

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Review of “Revealing Emerging Hydroclimatic Shifts: Advanced Trend Analysis of Rainfall and Streamflow in the Navasota River Watershed.”

Hydrology 4007227

This paper presents analyses of certain aspects of flooding dynamics in the NRW including precipitation trends and streamflow trends, as well as exploring the influence of autocorrelation on flooding probabilities, in general.  The paper is quite well-written.  The results, though modest, are relevant and could be valuable to local and regional water resource managers.  I have several minor comments/suggestions, and I list each individually below.  I think the paper could be accepted after minor revision.

Minor comments:

@ lines 176-177: the authors say that “O[bar] is the bar sign over the O.”  That tells too little.  I believe they mean that “O[bar] is the mean observed precipitation.”  In the same section, we are told, separately, that “P is PRISM rainfall,” and that “P denotes observed and PRISM rainfall mean….” (emphasis added).  Again, I think the authors mean that “P represents PRISM rainfall, and P[bar] represents the mean PRISM rainfall.”  I could easily be wrong in my interpretation, but that’s exactly the problem I have with these statements.  I recommend the authors revise to make perfectly clear what each variable represents.

The Paragraph beginning @ line 326, describing Figure 4, doesn’t match Figure 4.  The maps in Figure 4 show me that the northern region of the NRW shows marked increases when comparing early to later periods, and central and southern regions show modest or no change.  That is directly in contrast to the body of the paper, which describes these spatial outcomes in the opposite way.

The Paragraph starting @ line 369 is describing information in Table 2, but the text says “Mann-Kendall test,” while Table 2 shows Kendall-Tau results.  I assume this is merely a typo, but the text should match the table.  This could be related to the authors' frequent reference to the Kendall-Tau statistic as being one of the outcomes of the Mann-Kendall test.  But the MK test produces the test statistic M[sub]z, not Kendall's Tau.  These are separate tests, with separate and distinct results.  The authors could clarify this, especially because the MK test statistic is usually plotted as a time series for purposes of identifying the timing of significance of a trend, while Kendall's Tau provides info on the series in total (not as a time series).  I'd like to see this clarified for the reader.

Figure 7: I have a problem with the legend labeling on these figures.  Panels 7a and 7b are labeled identically for the main data plot (“yearly streamflow”).  But the panels display different data (7b is in fact yearly peak streamflow).  Also, 7c is mislabeled, as it depicts the 10-yr filtered annual peak streamflow.

Section 3.6 and Figure 8:  This is just my view, but when the results of a test are not statistically significant in any aspect, I suggest leaving it out entirely (which would mean removing it from the objectives and the Methods as well).  These two paragraphs and Figure 8 really don’t add any information.  Further, this result is not fundamentally necessary to your question; given that there was no regime shift centered on the year of dam construction, any other break point would have raised more questions than it answered (regarding the dam’s influence, at least).  I think that is an additional reason to exclude it from the paper.  I want to clarify the distinction I observe between this test and your earlier ones.  In each of your earlier tests there were insignificant results, but at least one aspect (one variable, say) came out significant.  That must be reported, of course, and thus it makes sense to include the directly related—but insignificant—results for those earlier tests.  However, on this test, with zero significant component to the results, I don’t think mentioning it is helpful, and worse, it could be distracting to the reader.

Section 4.4 can also be deleted, in accordance with the removal of the change detection analyses and results.

Author Response

REVIEWER 3

 RESPONSE TO THE REVIEW of the manuscript:

 Revealing Emerging Hydroclimatic Shifts: Advanced Trend Analysis of Rainfall and Streamflow in the Navasota River Watershed

Manuscript ID: hydrology-4007227

Dear Reviewer

Thank you for your time and effort in reviewing the manuscript. The comments were extremely helpful in improving the manuscript. Below we have repeated the comments and immediately following it, both our response and the improved text in the revised manuscript. The revised text is in blue.

 

Thanks again for your time and effort.

General comment

This paper presents analyses of certain aspects of flooding dynamics in the NRW including precipitation trends and streamflow trends, as well as exploring the influence of autocorrelation on flooding probabilities, in general.  The paper is quite well-written.  The results, though modest, are relevant and could be valuable to local and regional water resource managers.  I have several minor comments/suggestions, and I list each individually below.  I think the paper could be accepted after minor revision.

Response: Thank you for your thoughtful and constructive general comment. We appreciate your positive assessment of the manuscript and are pleased that you find the analysis relevant and potentially valuable for local and regional water resource management efforts. We also thank you for recognizing the clarity of the writing and the contribution of the results, even though they are modest in scope.

We have carefully considered each of your minor comments and suggestions and have revised the manuscript accordingly to improve clarity, accuracy, and overall presentation. We believe these revisions have strengthened the manuscript and addressed all concerns raised. We appreciate your recommendation for acceptance after minor revision and thank you for your valuable feedback throughout the review process.

 

Minor comments

Comment 1: @ lines 176-177: the authors say that “O[bar] is the bar sign over the O.”  That tells too little.  I believe they mean that “O[bar] is the mean observed precipitation.”  In the same section, we are told, separately, that “P is PRISM rainfall,” and that “P denotes observed and PRISM rainfall mean….” (emphasis added).  Again, I think the authors mean that “P represents PRISM rainfall, and P[bar] represents the mean PRISM rainfall.”  I could easily be wrong in my interpretation, but that’s exactly the problem I have with these statements.  I recommend the authors revise to make perfectly clear what each variable represents.

Response: Thank you for pointing out this ambiguity. We agree that the definitions of the variables were not sufficiently clear in the original text. In the revised manuscript, we have clarified the notation by explicitly defining each variable. Kindly see the revised equations and symbols of definition below.

Comment 2: The Paragraph beginning @ line 326, describing Figure 4, doesn’t match Figure 4.  The maps in Figure 4 show me that the northern region of the NRW shows marked increases when comparing early to later periods, and central and southern regions show modest or no change.  That is directly in contrast to the body of the paper, which describes these spatial outcomes in the opposite way.

Response: Thank you for identifying this important inconsistency. After reviewing the text and Figure 4, we agree that the original description did not accurately reflect the spatial patterns shown in the maps. We have revised the paragraph to ensure that the narrative accurately matches the spatial trends depicted in Figure 4. Kindly see our revision following the response to comment 6.

Comment 3: The Paragraph starting @ line 369 is describing information in Table 2, but the text says “Mann-Kendall test,” while Table 2 shows Kendall-Tau results.  I assume this is merely a typo, but the text should match the table.  This could be related to the authors' frequent reference to the Kendall-Tau statistic as being one of the outcomes of the Mann-Kendall test.  But the MK test produces the test statistic M[sub]z, not Kendall's Tau.  These are separate tests, with separate and distinct results.  The authors could clarify this, especially because the MK test statistic is usually plotted as a time series for purposes of identifying the timing of significance of a trend, while Kendall's Tau provides info on the series in total (not as a time series).  I'd like to see this clarified for the reader.

Response: Thank you for highlighting this important point. As you noted, the Mann–Kendall test produces the standardized test statistic Z_MK, which is used to assess the statistical significance of a monotonic trend, whereas Kendall’s Tau (τ) is a separate rank-based correlation coefficient that quantifies the strength and direction of that trend across the entire time series. In other words, τ is not the MK test statistic, but it is commonly reported alongside MK results to provide additional interpretive context regarding trend magnitude. To ensure clarity for readers, we have added a paragraph in the Methods section formally distinguishing between Z_MK and τ. Kindly see the added paragraph below.

Comment 4: Figure 7: I have a problem with the legend labeling on these figures.  Panels 7a and 7b are labeled identically for the main data plot (“yearly streamflow”).  But the panels display different data (7b is in fact yearly peak streamflow).  Also, 7c is mislabeled, as it depicts the 10-yr filtered annual peak streamflow.

 

Response: Thank you for pointing out the labeling issues in Figure 7. We agree that the original legends and panel labels did not clearly distinguish among the different streamflow variables. We have revised the figure accordingly to ensure each panel is accurately and clearly labeled. Kindly refer to the updated version of Figure 7 below.

Comment 5: Section 3.6 and Figure 8:  This is just my view, but when the results of a test are not statistically significant in any aspect, I suggest leaving it out entirely (which would mean removing it from the objectives and the Methods as well).  These two paragraphs and Figure 8 really don’t add any information.  Further, this result is not fundamentally necessary to your question; given that there was no regime shift centered on the year of dam construction, any other break point would have raised more questions than it answered (regarding the dam’s influence, at least).  I think that is an additional reason to exclude it from the paper.  I want to clarify the distinction I observe between this test and your earlier ones.  In each of your earlier tests there were insignificant results, but at least one aspect (one variable, say) came out significant.  That must be reported, of course, and thus it makes sense to include the directly related—but insignificant—results for those earlier tests.  However, on this test, with zero significant component to the results, I don’t think mentioning it is helpful, and worse, it could be distracting to the reader.

Response: Thank you for the comment. After revising our Pettitt test analysis using updated Python code based on comments of other reviewers, we identified a statistically significant hydrologic shift around 1990 in the 10-year moving average streamflow. This result aligns with the rainfall changes reported in Section 3.2 and the post-dam hydrologic patterns discussed in Section 4.1, resolving the earlier inconsistency. Sections 3.2, 4.1, and 4.4 have been updated accordingly. Please review our revisions of the sections following the response of comment 6.

Comment 6: Section 4.4 can also be deleted, in accordance with the removal of the change detection analyses and results.

Response: Thank you for the suggestion. We have retained Section 4.4 because the revised Pettitt test analysis now identifies a statistically significant hydrologic shift around 1990 in the 10-year moving average streamflow. This result supports the relevance of the change-detection discussion and aligns with the rainfall and streamflow patterns presented in Sections 3.2 and 4.1. Therefore, Section 4.4 remains an important component of the updated analysis. Kindly see our revision for section 4.4 below.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I want to thank the authors for their efforts in revising the manuscript. I recommend that the paper be accepted.