the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 03 Feb 2025
Seasonal Predictability of Vapor Pressure Deficit in the western United States
Abstract. Saturation vapor pressure deficit (VPD), a measure of the difference between how much moisture the atmosphere can hold versus how much is present, is highly correlated with the annual mean area burned by wildland fires in the western United States. The present analysis uses linear inverse models (LIMs) to forecast seasonal VPD and decompose skill into contributions from a nonlinear trend, coupled sea surface temperature (SST)-VPD variability, and VPD-only variability. Subregions of the western US are considered using Geographic Area Coordination Centers which have different times of year and lead times for which VPD forecast skill is greatest. However, the sources of skill are similar among the subregions. In LIM forecasts, particularly those made for summer and early fall, the trend contributes to VPD skill up to 18 months in advance, with a secondary contribution from internal VPD variability at lead times of one to two months. Positive SST and VPD anomalies and negative soil moisture anomalies are associated with the positive sign of the trend time series, which has been observed without interruption since the late 1990s. Coupled SST-VPD variability contributes to VPD skill mainly for forecasts verifying between December to May for lead times up to 12 months in some subregions. Forecasts that are especially skillful and display high confidence, seasonal forecasts of opportunity (SFOs), are associated with SSTs that produce high VPD skill over California, the Southwest, and Texas, while internal VPD anomalies contribute to skill over the Great Basin and western Northern Plains. SFOs are initialized during periods of El Niño-Southern Oscillation development, with La Niña SSTs associated with positive western US VPD anomalies and consequently, enhanced wildland fire risk.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2456 KB)
-
Supplement
(740 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2456 KB) - Metadata XML
-
Supplement
(740 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
- RC1: 'Comment on egusphere-2025-115', Anonymous Referee #1, 27 Mar 2025
-
RC2: 'Comment on egusphere-2025-115', Anonymous Referee #2, 28 May 2025
This study explores the seasonal predictability of vapor pressure deficit (VPD) in the western United States, a region increasingly impacted by wildfires. The use of Linear Inverse Models (LIMs) to forecast VPD and decompose contributions from a nonlinear trend, SST-VPD coupling, and internal VPD variability is both technically rigorous and conceptually insightful. The manuscript is well-organized, clearly written, and the inferences drawn from its results make generally good sense.
I recommend publication pending moderate revisions. Below are several specific comments for clarification and improvement:
- The authors use JRA-55 SST as input for the LIM, rather than more widely used SST datasets such as HadISST or ERSST, which are commonly validated in ENSO and SST variability studies. Given the important role of SST in the decomposition of VPD predictability, it would strengthen the manuscript to either justify this choice or test the sensitivity of results to other SST products.
- The authors calculate monthly VPD using monthly mean temperature and RH, which may underestimate extremes due to the nonlinear relationship between temperature and vapor pressure. While this method may be appropriate for seasonal-scale prediction, I think an explanation is needed for this choice over computing the monthly average of daily VPD, which is commonly used in fire-weather studies.
- While anomaly correlation coefficient (ACC) is a standard metric, it does not capture magnitude errors or bias. For example, what are the root mean square error (RMSE) values corresponding to Figures 1 and 2? Including such metrics would provide a more comprehensive assessment of model performance.
- It is not quite clear how the SST pattern in Figure 12a resembles the Pacific Meridional Mode (PMM). Could the authors consider quantifying this similarity, for example by providing a spatial correlation coefficient with a canonical PMM pattern?
- Could the limited detrended VPD skill in ASO be attributed to the weak influence of SST on large-scale atmospheric circulation over the US during this season? The authors might consider discussing other factors (e.g., Northern Hemisphere teleconnection or Arctic sea ice) here, which may also play a more important role.
References:
Zou, Y., Rasch, P. J., Wang, H., Xie, Z., & Zhang, R. (2021). Increasing large wildfires over the western United States linked to diminishing sea ice in the Arctic. Nature communications, 12(1), 6048.
Liu, S., Hu, S., & Seager, R. (2024). The West Pacific teleconnection drives the interannual variability of autumn wildfire weather in the western United States after 2000. Earth's Future, 12(11), e2024EF004922.
Lou, J., Joh, Y., Delworth, T. L., & Jia, L. (2025). Identifying source of predictability for vapor pressure deficit variability in the southwestern United States. npj Climate and Atmospheric Science, 8(1), 139.
Citation: https://doi.org/10.5194/egusphere-2025-115-RC2 - AC1: 'Author Response to Reviewers on egusphere-2025-115', Melissa Breeden, 30 Jun 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-115', Anonymous Referee #1, 27 Mar 2025
General comments:
By using a linear inverse modeling framework for the fields Vapor Pressure Deficit (VPD), Sea Surface Temperature (SST), and Soil Moisture (SM), the authors forecast seasonal VPD for the western United States. The study identifies Seasonal Forecasts of Opportunity (SFOs) and links them to ENSO. The manuscript is well-written, with no significant flaws in the logic. The figures are clearly presented, and the results support the conclusions. I recommend publication of this manuscript with minor revisions. I would like the authors to consider the following comments.
Major comments:
The use of soil moisture (SM) needs more justification. What would the results look like if SM were removed from the analysis? If there is a significant difference between the forecasts with and without SM, what exactly is SM capturing?
The discussion on ENSO using the leading optimal patterns was succinct, but would it be more valuable for a general audience to include a direct analysis linking SFOs to ENSO indices? For example, similar to Breeden et al. (2022) (their Figs. 6 & 7), authors could overlay the SFOs on the Niño 3.4 index or the Oceanic Niño Index.
Specific comments:
L68: Beverly et al (2022) is missing from the reference list.
L311: Mariotti et al (2020) is missing from the reference list.
L399-401: The first two sentences about winds in this paragraph seem out of place, especially since the next sentence mentions other variables, including winds. Consider revising for clarity and consistency to ensure a smoother flow of ideas.
Â
References:
Breeden, M.L., Albers, J.R. and Hoell A.: Subseasonal precipitation forecasts of opportunity over southwest Asia. Weather and Climate Dynamics, 3, 1183–1197, https://doi.org/10.5194/wcd-3-1183-2022, 2022.
Citation: https://doi.org/10.5194/egusphere-2025-115-RC1 -
RC2: 'Comment on egusphere-2025-115', Anonymous Referee #2, 28 May 2025
This study explores the seasonal predictability of vapor pressure deficit (VPD) in the western United States, a region increasingly impacted by wildfires. The use of Linear Inverse Models (LIMs) to forecast VPD and decompose contributions from a nonlinear trend, SST-VPD coupling, and internal VPD variability is both technically rigorous and conceptually insightful. The manuscript is well-organized, clearly written, and the inferences drawn from its results make generally good sense.
I recommend publication pending moderate revisions. Below are several specific comments for clarification and improvement:
- The authors use JRA-55 SST as input for the LIM, rather than more widely used SST datasets such as HadISST or ERSST, which are commonly validated in ENSO and SST variability studies. Given the important role of SST in the decomposition of VPD predictability, it would strengthen the manuscript to either justify this choice or test the sensitivity of results to other SST products.
- The authors calculate monthly VPD using monthly mean temperature and RH, which may underestimate extremes due to the nonlinear relationship between temperature and vapor pressure. While this method may be appropriate for seasonal-scale prediction, I think an explanation is needed for this choice over computing the monthly average of daily VPD, which is commonly used in fire-weather studies.
- While anomaly correlation coefficient (ACC) is a standard metric, it does not capture magnitude errors or bias. For example, what are the root mean square error (RMSE) values corresponding to Figures 1 and 2? Including such metrics would provide a more comprehensive assessment of model performance.
- It is not quite clear how the SST pattern in Figure 12a resembles the Pacific Meridional Mode (PMM). Could the authors consider quantifying this similarity, for example by providing a spatial correlation coefficient with a canonical PMM pattern?
- Could the limited detrended VPD skill in ASO be attributed to the weak influence of SST on large-scale atmospheric circulation over the US during this season? The authors might consider discussing other factors (e.g., Northern Hemisphere teleconnection or Arctic sea ice) here, which may also play a more important role.
References:
Zou, Y., Rasch, P. J., Wang, H., Xie, Z., & Zhang, R. (2021). Increasing large wildfires over the western United States linked to diminishing sea ice in the Arctic. Nature communications, 12(1), 6048.
Liu, S., Hu, S., & Seager, R. (2024). The West Pacific teleconnection drives the interannual variability of autumn wildfire weather in the western United States after 2000. Earth's Future, 12(11), e2024EF004922.
Lou, J., Joh, Y., Delworth, T. L., & Jia, L. (2025). Identifying source of predictability for vapor pressure deficit variability in the southwestern United States. npj Climate and Atmospheric Science, 8(1), 139.
Citation: https://doi.org/10.5194/egusphere-2025-115-RC2 - AC1: 'Author Response to Reviewers on egusphere-2025-115', Melissa Breeden, 30 Jun 2025
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 676 | 111 | 35 | 822 | 50 | 42 | 45 |
- HTML: 676
- PDF: 111
- XML: 35
- Total: 822
- Supplement: 50
- BibTeX: 42
- EndNote: 45
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Melissa Leah Breeden
Andrew Hoell
Rochelle Pauline Worsnop
John Robert Albers
Michael T. Hobbins
Rachel Maya Robinson
Daniel James Vimont
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2456 KB) - Metadata XML
-
Supplement
(740 KB) - BibTeX
- EndNote
- Final revised paper
General comments:
By using a linear inverse modeling framework for the fields Vapor Pressure Deficit (VPD), Sea Surface Temperature (SST), and Soil Moisture (SM), the authors forecast seasonal VPD for the western United States. The study identifies Seasonal Forecasts of Opportunity (SFOs) and links them to ENSO. The manuscript is well-written, with no significant flaws in the logic. The figures are clearly presented, and the results support the conclusions. I recommend publication of this manuscript with minor revisions. I would like the authors to consider the following comments.
Major comments:
The use of soil moisture (SM) needs more justification. What would the results look like if SM were removed from the analysis? If there is a significant difference between the forecasts with and without SM, what exactly is SM capturing?
The discussion on ENSO using the leading optimal patterns was succinct, but would it be more valuable for a general audience to include a direct analysis linking SFOs to ENSO indices? For example, similar to Breeden et al. (2022) (their Figs. 6 & 7), authors could overlay the SFOs on the Niño 3.4 index or the Oceanic Niño Index.
Specific comments:
L68: Beverly et al (2022) is missing from the reference list.
L311: Mariotti et al (2020) is missing from the reference list.
L399-401: The first two sentences about winds in this paragraph seem out of place, especially since the next sentence mentions other variables, including winds. Consider revising for clarity and consistency to ensure a smoother flow of ideas.
Â
References:
Breeden, M.L., Albers, J.R. and Hoell A.: Subseasonal precipitation forecasts of opportunity over southwest Asia. Weather and Climate Dynamics, 3, 1183–1197, https://doi.org/10.5194/wcd-3-1183-2022, 2022.