Reitz (2026) Quantifying and interpreting annual and daily slopes between actual and potential evapotranspiration by land cover type

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-10-31
• Authors: Meredith D. Reitz
• DOI: 10.1016/j.jhydrol.2026.136711

Research Groups

U.S. Geological Survey, Hydrologic Remote Sensing Branch

Short Summary

This study quantifies annual and daily relationships between actual (ET) and potential (PET) evapotranspiration across the conterminous U.S. by land cover type, partially controlling for radiative forcing. It finds negative correlations annually, consistent with complementary theory, but highly variable daily ET-PET slopes that depend on time of year and land cover, which are not consistently explained by either demand-driven or complementary frameworks, suggesting active plant physiological moderation.

Objective

• To test competing conceptual frameworks (demand-driven vs. complementary theory) for the relationship between ET and PET by partially controlling for the influence of radiative forcing.
• To quantify whether the relationship between ET and PET shows a positive or negative correlation at annual and daily timescales.

Study Configuration

• Spatial Scale: Conterminous U.S. (1,816 watersheds and 159 AmeriFlux flux tower sites).
• Temporal Scale: Annual (79,959 water-years from 1896 to 2018) and Daily (174,989 days from 1991 to 2018).

Methodology and Data

• Models used: 13 established Potential Evapotranspiration (PET) equations (e.g., Priestley-Taylor, Milly-Dunne, Blaney-Criddle, Eagleman, Hamon, Schendel, Baier-Robertson, FoPM, Hargreaves, FAO Penman-Monteith, FAO24, Jensen-Haise, Makkink).
• Data sources:
  • Water balance ET: U.S. Geological Survey (USGS) National Water Information System (outflow), Parameter-Elevation Regressions on Independent Slopes Model (PRISM) (precipitation), USGS water use data (groundwater irrigation adjustment).
  • Flux tower ET: AmeriFlux database.
  • Climatic variables (temperature, vapor pressure deficit): PRISM daily 4-km product.
  • Wind speed: gridMET model (monthly averages).
  • Radiation (daylight hours, extraterrestrial, solar, net radiation): Calculated using established equations.
  • Land cover: USGS Earth Resources Observation and Science (EROS) Center (1938–1999) and 2016 National Land Cover Database (NLCD) (2000–2018).
  • Koppen climate zones: AmeriFlux Biological, Ancillary, Disturbance and Metadata (BADM) files.

Main Results

• At the annual timescale, negative slope values were observed between ET and PET across all land cover types and PET equations, consistent with the atmospheric feedback described by complementary theory.
• At the daily timescale, the sign and magnitude of the ET-PET slope varied strongly with the time of year and land cover type, not consistently explained by either demand-driven or complementary theory.
  • PET equations dependent on energy only showed generally positive correlations, with no statistically significant differences in ET-PET slope among vegetation types (F(4) = 1.32, p = 0.26).
  • PET equations dependent on atmospheric variables only showed some negative correlations, with drier land covers tending to exhibit more negative slopes. Differences in ET-PET slope among vegetation types were statistically significant (F(4) = 6.66, p < 0.001).
  • PET equations dependent on both energy and atmospheric variables showed the clearest stratification by land cover type, with higher annual precipitation correlating with more positive slopes and lower annual precipitation with more negative slopes. Both positive and negative slopes reached their greatest magnitudes in the summer. Differences were statistically significant (F(4) = 26.22, p < 0.001).
• Calculated 'b' values from the complementary theory framework were frequently negative (46% of cases), indicating observed ET in excess of PET or computed PET lower than wet-environment ET (PETw), which is not well explained by the theory.
• Neglecting these dynamic, landscape-specific ET-PET relationships could result in daily ET estimation errors ranging from 32% to 50% of the average ET value.
• No statistically significant correlation was found between the temporal variation of ET-PET slopes and monthly precipitation (R2 = 0.005, p = 0.60), wind speed (R2 = 0.07, p = 0.038), or variation in net radiation (R2 = 0.0098, p = 0.45).

Contributions

• Provides a comprehensive quantification of annual and daily ET-PET slopes across diverse land cover types and a wide range of established PET equations for the conterminous U.S.
• Demonstrates that neither the demand-driven nor complementary conceptual frameworks consistently explain the observed daily ET-PET relationships when the first-order influence of radiative forcing is partially controlled.
• Offers a novel interpretation that the observed land cover-specific and time-varying daily ET-PET slopes are consistent with active plant physiological moderation and adaptation to typical water availability conditions.
• Highlights a significant opportunity to improve the accuracy of current ET estimation methods by incorporating these dynamic, landscape-specific dependencies of ET on PET.
• Critiques the endogeneity of PET in the Budyko framework, suggesting that using only exogenous, radiation-based control variables would be more statistically defensible.

Funding

• U.S. Department of Energy Office of Science, Office of Biological and Environmental Research (supported the AmeriFlux Management Project under contract number DE-AC02-05CH11231).

Citation

@article{Reitz2026Quantifying,
  author = {Reitz, Meredith D.},
  title = {Quantifying and interpreting annual and daily slopes between actual and potential evapotranspiration by land cover type},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136711},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136711}
}