the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Temporal persistence of postfire flood hazards under present and future climate conditions in southern Arizona, USA
Abstract. Changes to soil hydraulic properties that reduce infiltration capacity following fire can increase flash flood risks. These risks are exacerbated by rainfall intensification associated with a warming climate. However, the potential effects of climate-change-driven rainfall intensification on postfire floods remain largely unexplored. Using rainfall and runoff observations from a 49.4 km2 watershed in southern Arizona, USA, and a hydrological model (KINEROS2), we examined the temporal evolution following a historic fire of three crucial hydrologic parameters: soil saturated hydraulic conductivity (Ksp), net capillary drive (Gp), and hydraulic roughness (nc). We explored how the effect of fire on these parameters may influence peak flow under future climate scenarios derived from CMIP6, specifically the medium emissions scenario (SSP245) and high emissions scenario (SSP585). Results demonstrate an increase in Ksp from 11 mm/hr in the first postfire year to 60 mm/hr in postfire year three. Gp similarly increased from 19 mm in the first postfire year to 30 mm in the third, while nc was relatively constant. The highest simulated Qp occurred in the first postfire year. Under the SSP245 scenario, the likelihood of a 100-yr flood is projected to be twice as large by the middle century relative to its historical magnitude. Simulations further indicate that the maximum expected discharge associated with a postfire flood, as derived from historical data, could be triggered by a 10-yr rainstorm under the SSP585 scenario by the late century. Simulations also demonstrate that rainfall intensification will lead to greater persistence of elevated flood hazards following fire by late century under both the SSP245 and SSP585 scenarios.
- Preprint
(2059 KB) - Metadata XML
-
Supplement
(509 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on nhess-2024-151', Anonymous Referee #1, 11 Feb 2025
The manuscript authored by Liu et al. addresses a relevant scientific topic in the framework of natural hazards, such as the hydrological response of burned watersheds under present and future climatic conditions. The analyzed case study is in the USA, where several research teams are working on the same hazard due to relevant impacts associated to post-fire floods and debris flows occurring every year. The work is well presented and described, and it could be helpful to a broad community focusing on fire-related hazards and effects of climate change.
After a careful revision, I have identified a series of points that the Authors should address before considering the manuscript ready for publication. As general remarks, the Authors should clarify that they worked on a large watershed not prone to quick flooding response like those affected by post-fire flash floods that, however, respond to sub-hourly rainfall. In addition, I did not find field measurements performed in the three years following the wildfire to validate outputs of the used Kineros 2 model, in terms of soil saturated hydraulic conductivity, net capillary drive, and hydraulic roughness. Some of the model outputs support the Authors in the definition of the window of disturbance associated to the analyzed site, that I suggest to stress together with the distinction between rainfall-runoff events strictly related to the wildfire and those that can be considered as ordinary river floods (i.e., not fire-related). The simulated increasing of the peak discharges should be also discussed in terms of liner or non-linear relationships with rainfall intensification.
For these and other minor comments included in the attached pdf file, I suggest minor revision.
-
RC2: 'Comment on nhess-2024-151', Anonymous Referee #2, 17 Feb 2025
The main objectives of this study were to calibrate KINERO2 (K2) hydrological model and to use the hydrological parameters obtained to simulate discharges under present and future climate conditions. The study uses observed rainfall and runoff data collected in a watershed located in Arizona and burned in 2020. Model predictions under different climate scenarios are relevant to stakeholders whose knowledge of potential runoff is crucial for this area, which is subject to short-duration, high-intensity rainfall events. The text is clear and easy to read. After revision, the paper is ready to be published in Natural Hazards and Earth System Sciences journal.
Line 149 – Explain how the runoff and baseflow were separated from the discharge. If this was not done, explain the approach used.
Line 153 – Why were Pre-4 and Pre-10 events selected?
Line 176 – I assume SBS stands for soil burn severity, but it is never mentioned in the text. It will be easier for the reader if you spell it down.
Line 177 – 179 – From what I understand the defaults in the AGWA tables are lower than those calibrated for burned areas. Isn't this interesting? Why does this happen?
Line 241 – What is ARI? Can you describe it?
Line 285 – The model reproduced the observed runoff response reasonably well. How did you state that for pre-4 and pre-10 if you don't have the KGE values (Figure 4)?
Line 303 – 307 – shouldn’t this part go in the methods?
Line 395 – It is clear from the results that KINEROS2 model better simulates infiltration-excess runoff but it is unable to simulate other runoff-generation mechanisms. You mentioned ParFlow model. Why did you choose the KINEROS2 model and not in a model able to simulate a mixture of flow dynamics? It is because the goal was to simulate infiltration-excess conditions and, in this way, reduce the range of hydrological parameters values, thus improving predictions? Maybe stress out between line 445-458 that similar simulations settings will be for events driven by infiltration excess runoff.
Line 409 – Do you have KGE values for pre-4 and pre-10 events? I would show them because otherwise how can you infer that the Ksp is greater than 60 mm/hour in the third post-fire year based on a single event?
Line 445-448 - Perhaps you could point out that similar simulations will be set up for events driven by infiltration-excess runoff.
Figure 7 - Why did you not show the ARI = 10 years?
Citation: https://doi.org/10.5194/nhess-2024-151-RC2
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
274 | 51 | 12 | 337 | 21 | 8 | 14 |
- HTML: 274
- PDF: 51
- XML: 12
- Total: 337
- Supplement: 21
- BibTeX: 8
- EndNote: 14
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1