the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Nov 2025
Improved model predictions of carbon and water fluxes by including drought legacy effects
Abstract. Besides simultaneous influences, droughts have lasting impacts on vegetation by impairing hydraulic and photosynthetic capacities, known as the drought legacy effects. The ignorance of legacy effects in numerical simulations, such as lagged xylem recovery, may lead to significant model-observation discrepancies. However, the limited temporal resolution of most observational data makes it challenging to capture the physiological dynamics necessary to improve model accuracy. Here, we investigated the recovery of carbon flux (represented by gross primary productivity, GPP) and water flux (represented by evapotranspiration, ET) following a severe drought in 2012, using half-hourly eddy-covariance flux observations and weekly predawn leaf water potential measurements from a temperate forest in the Central US. We implemented both optimality-based and empirical stomatal models within a land surface model, testing three drought recovery scenarios for each: no recovery, full recovery, and partial recovery of xylem hydraulic conductance and photosynthetic capacity. Before and during the drought, all stomatal models performed similarly for GPP and ET. Post-drought, assuming no recovery led to underestimated ET; assuming full recovery led to overestimated GPP; and assuming partial recovery improved both, indicating persistent biochemical limitations after drought. The observed carbon-water decoupling during and after the event further points to non-stomatal constraints on photosynthesis and unequal stress on carbon and water fluxes. Our work highlights the need to account for delayed recovery of xylem hydraulics and photosynthetic capacity when modeling drought legacy effects. Further research to mechanistically represent dynamic recovery processes, particularly their timing and magnitude, is essential for improving the modeling of global carbon and water fluxes.
- Preprint
(1181 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-5684', Anonymous Referee #1, 18 Dec 2025
-
RC2: 'Comment on egusphere-2025-5684', Xin Yu, 05 Jan 2026
In this manuscript, Yao et al. examine drought legacy effects on carbon and water fluxes using eddy-covariance observations and a process-based model for a temperate forest in the central US. The manuscript is generally well written and reflects considerable effort in the model experiments and data analysis. It has the potential to contribute to the literature by providing a data–model fusion framework for understanding drought legacy effects and improving the representation of post-drought recovery of carbon and water fluxes. I only have several concerns that should be addressed before publication.
Major comments:
1) The overall structure of the methodology is understandable, but some key details are missing. The equations for how GPP, ET, leaf water potential, and hydraulic conductance are calculated should be included. They are important for readers to understand how the water stress factor affects the dynamics of these variables.
2) The authors show the comparison of GPP and ET between observations and model simulations under different scenarios across drought stages in an aggregated way. It would be great to show the temporal dynamics in more detail by showing the daily time series, which could help convey the messages more clearly to readers. In addition, the time series of the water stress and hydraulic conductance under the partial recovery scenario could be worth showing to demonstrate the role of the partial recovery of hydraulic conductance in the recovery of carbon and water fluxes.
3) Because the model does not simulate soil evaporation and re-evaporation of rainfall interception, the authors assume a fixed T/ET ratio to link transpiration (T) to observed ET. However, this is a strong assumption in the context of drought legacy effects on ET. Because, given the same conditions of water and energy availability, the reduced T due to legacy effects could be compensated by the soil evaporation, resulting in a similar magnitude of ET. Therefore, the result that the ET simulation under the full-recovery assumption is closer to the observation than the non-recovery assumption could be confounded by this compensation effect. This could also question the finding that a quicker resumption of ET, but a slower recovery of GPP. Given that the model used can not simulate soil evaporation, I suggest authors try to use an ET partitioning method (e.g., Nelson et al., 2020) to estimate transpiration and directly compare it with model simulations.
4) I am a little confused about the true mechanisms investigated in this study. The authors design three hydraulic conductance recovery scenarios and use a water stress factor (β), which gives the impression that the effect of partial recovery of hydraulic conductance is examined. However, in the analysis, this β factor is used to adjust Vcmax to account for the down-regulation of photosynthetic capacity, and in the discussion later, the significance of Vcmax impairment in shaping the delayed recovery of carbon fluxes is highlighted. Is the down-regulation of photosynthetic capacity the consequence of the loss of hydraulic conductance? Or they jointly contribute to drought legacy effects. There seems to be a mixture of these two processes. It would be great if this point could be clarified throughout the manuscript.
Line-by-line comments:
Line 4: repeated numbers
Line 23: timing? You mean duration?
Line 35: From my understanding, Ciais et al. (2005) did not include legacy effects. Also, it is unclear if drought legacy effects could shift ecosystems from carbon sinks to carbon sources. Maybe consider rephrasing this sentence.
Line 149: Where is Figure S1? The entire supplement seems to be missing.
Lines 172-173: It would be great if the equation of the optimality stomatal model or more details could be provided. How to apply β to it?
Table 1: Where is Vcmax?
Line 229: just curious. Why 63% instead of 50%?
Figure 1b): would be great to show the related equations to better understand this figure. Also see the first major comment.
Figure 4: Why is the difference between RMSEFR in b and c so large? They should be based on the same model without β, right?
Lines 384-390: see the 3rd major comment.
Lines 484-485: see the 3rd major comment.
Lines 488-489: From my understanding, under the partial recovery scenario, the model needs to be calibrated by observations during the recovery phase. Would this prevent it from applying to the case where the observation is not available?
Citation: https://doi.org/10.5194/egusphere-2025-5684-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 287 | 82 | 23 | 392 | 13 | 15 |
- HTML: 287
- PDF: 82
- XML: 23
- Total: 392
- BibTeX: 13
- EndNote: 15
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript presents a modelling study using the CliMA land surface model investigating the effect of assumptions around tree recovery after drought at one site which experienced a severe drought in 2012. The results show that including a mechanism for partial rather than full recovery leads to a better fit to data in the post-drought period. This is an important topic as models still frequently assume instantaneous responses to environmental conditions and there is no standard framework for representing lagged responses.
In general, the analysis is scientifically sound but the paper itself need some edits to make it more robust and easier to understand.
Major comments
Minor comments
L 135 does the MODIS LAI respond to drought at the site? Does having this as an input partially represent the effects of drought already?
L 150 unclear why, given the research questions, it is necessary to evaluate canopy spectra
L 167 “Changing this T/ET ratio does not impact our results.” Some evidence is needed here
L 167 “The CliMA Land model was employed to reproduce the dynamics of carbon and water fluxes during the drought period.” This statement feels superfluous here.
L 171 “several stomatal optimality models” sever feels very vague here
L 189 table 1 – this table has only 1 line, but above you say you also used an optimal model
Section 2.4 – many more details are needed here, preferably in the form of equations. It is not enough to say that vulnerability curves or the data assimilation method is taken from another paper
L 215 Table 2 – unclear what information this table add as everything is just a tick. The notations for the two options for the Medlyn model also do not match those in Table 1
L 231 what is Kmax? Seems like an important parameter but it is the first time it’s mentioned I think
L 257 is this the same method as the data assimilation for the partial recovery scenario above or something else?
L 335 But the model overestimates the coupling during the wilted period as well, so the issue here is not just the full recovery assumption
L 338 this looks like a repetition of the methods here
L 346 it might help here to show a timeseries of model GPP and ET to understand the evolution of recovery
L 421 it feels odd to mention simulations at other sites in the discussion for the first tiem when this was not mentioned in the methods