A process-based representation of hydraulic activation of stomata induced by hygroscopic fertilizer salts in the ORCHIDEE-CROP model (v2.2)

Abstract. Hydraulic activation of stomata (HAS) induced by hygroscopic salt films on leaf surfaces can create an additional liquid water loss pathway beyond conventional stomatal vapor diffusion. However, this process is not represented in current land surface and crop models. Using the practice-relevant example of hygroscopic urea ammonium nitrate (UAN) application to field crops, we developed a process-based parameterization of UAN-induced HAS in the ORCHIDEE-CROP model. The scheme represents event-based UAN retention, prognostic HAS storage, dry depletion, rainfall wash-off, and a relative humidity-dependent effective wet fraction, which is incorporated into the latent heat flux (LE) partitioning framework through an additional evaporative efficiency term. The parameterization was calibrated using event-scale transpiration anomalies from a winter wheat site and evaluated at three winter wheat sites and one maize site using eddy covariance (EC) observations and crop management records. The scheme produced transient post-UAN HAS responses, with declining HAS storage and event-specific wetting dynamics. Including HAS increased simulated mean LE during the 0–10 d post-UAN period by 2.0–4.0 % across the four study sites, while soil water feedback reduced plant transpiration, partially offsetting the direct HAS water loss, with mean offset fractions ranging from 13.3 % to 24.9 % across sites. Comparison with EC observations showed that including HAS tended to reduce negative LE bias when the standard model underestimated observed LE. Simulated gross primary productivity remained nearly unchanged, whereas increased LE reduced simulated inherent water-use efficiency by 4.0–5.3 % across the four study sites. These results demonstrate the feasibility of representing UAN-induced HAS as an event-driven, meteorologically regulated process and provide a basis for extending the framework to hygroscopic aerosol effects in land surface and crop models.