Vegetation optimality explains the convergence of catchments on the Budyko curve
<p>The Budyko framework puts the long-term mean annual evapotranspiration (ET) of a catchment in relation to its maximum possible value determined by the conservation of mass (ET cannot exceed mean annual precipitation) and energy (ET can not exceed mean annual net radiation) in the absence of...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2022-12-01
|
| Series: | Hydrology and Earth System Sciences |
| Online Access: | https://hess.copernicus.org/articles/26/6289/2022/hess-26-6289-2022.pdf |
| _version_ | 1811200205981220864 |
|---|---|
| author | R. C. Nijzink S. J. Schymanski |
| author_facet | R. C. Nijzink S. J. Schymanski |
| author_sort | R. C. Nijzink |
| collection | DOAJ |
| description | <p>The Budyko framework puts the long-term mean annual evapotranspiration (ET) of a catchment in relation to its maximum possible value determined by the conservation of mass (ET cannot exceed mean annual precipitation) and energy (ET can not exceed mean annual net radiation) in the absence of significant storage contributions. Most catchments plot relatively close to this physical limit, which allowed the development of an empirical equation (often referred to as the Budyko curve) for estimating mean annual evaporation and runoff from observed net radiation and precipitation. Parametric forms of the curve often use a shape parameter, <span class="inline-formula"><i>n</i></span>, that is seen as a catchment characteristic. However, a satisfying explanation for the convergence and self-organization of catchments around such an empirical curve is still lacking. In this study, we explore if vegetation optimality can explain the convergence of catchments along a Budyko curve and in how far can <span class="inline-formula"><i>n</i></span> be seen as a catchment characteristic.</p>
<p>The Vegetation Optimality Model (VOM) optimizes vegetation properties and behavior (e.g., rooting depths, vegetation cover, stomatal control) to maximize the difference between the total carbon taken up from the atmosphere and the carbon used for maintenance of plant tissues involved in its uptake, i.e., the long-term net carbon profit (NCP). This optimization is entirely independent of observed ET and hence the VOM does not require calibration for predicting ET. In a first step, the VOM was fully optimized for the observed atmospheric forcing at five flux tower sites along the North Australian Tropical Transect, as well as 36 additional locations near the transect and six Australian catchments. In addition, the VOM was run without vegetation for all sites, meaning that all precipitation was partitioned into soil evaporation and runoff. For comparison, three conceptual hydrological models (TUWmodel, GR4J, and FLEX) were calibrated for the Australian catchments using the observed precipitation and runoff. Subsequently, we emulated step changes in climate by multiplying precipitation (<span class="inline-formula"><i>P</i></span>) by factors ranging between 0.2 and 2 before running the VOM and hydrological models without changing the vegetation properties or model parameters, emulating invariant catchment characteristics under a changed climate. In a last step, the VOM was re-optimized for the different <span class="inline-formula"><i>P</i></span> amounts, allowing vegetation to adapt to the new situation. Eventually, Budyko curves were fit by adapting the parameter <span class="inline-formula"><i>n</i></span> to the model results. This was carried out for both multiple sites simultaneously and for each individual study site, thereby assuming that <span class="inline-formula"><i>n</i></span> is a site-specific characteristic.</p>
<p>The optimized VOM runs tracked relatively close to a Budyko curve with a realistic <span class="inline-formula"><i>n</i></span> value and close to observations, whereas the runs without vegetation led to significantly lower evaporative fractions and unrealistically low <span class="inline-formula"><i>n</i></span> values compared with literature. When fitting <span class="inline-formula"><i>n</i></span> to individual catchments, changes in <span class="inline-formula"><i>P</i></span> led to changes in <span class="inline-formula"><i>n</i></span> (increasing <span class="inline-formula"><i>n</i></span> for decreasing <span class="inline-formula"><i>P</i></span>) in all model runs (including the three conceptual models) except if the VOM was re-optimized for each change in <span class="inline-formula"><i>P</i></span>, which brought the value of <span class="inline-formula"><i>n</i></span> back close to its value for the unperturbed <span class="inline-formula"><i>P</i></span> in each catchment. For the re-optimized VOM runs, the variation in <span class="inline-formula"><i>n</i></span> between catchments was greater than within each catchment in response to multiplications of <span class="inline-formula"><i>P</i></span> with a factor 0.2 to 2.</p>
<p>These findings suggest that optimality may explain the self-organization of catchments in Budyko space, and that the accompanying parameter <span class="inline-formula"><i>n</i></span> does not remain constant for constant catchment and vegetation conditions as hypothesized in the literature, but in fact emerges through the adaptation of vegetation to climatic conditions in a given hydrological setting. Moreover, the results suggest that <span class="inline-formula"><i>n</i></span> might initially increase in response to suddenly reduced <span class="inline-formula"><i>P</i></span>, and only slowly returns to its original, catchment-specific value, as vegetation re-adjusts to the new climate over decades and centuries. This may constitute a new basis for the evaluation and prediction of catchment responses to climatic shifts.</p> |
| first_indexed | 2024-04-12T01:59:44Z |
| format | Article |
| id | doaj.art-e3e006c103694db0b45ab310d9d991a5 |
| institution | Directory Open Access Journal |
| issn | 1027-5606 1607-7938 |
| language | English |
| last_indexed | 2024-04-12T01:59:44Z |
| publishDate | 2022-12-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Hydrology and Earth System Sciences |
| spelling | doaj.art-e3e006c103694db0b45ab310d9d991a52022-12-22T03:52:42ZengCopernicus PublicationsHydrology and Earth System Sciences1027-56061607-79382022-12-01266289630910.5194/hess-26-6289-2022Vegetation optimality explains the convergence of catchments on the Budyko curveR. C. NijzinkS. J. Schymanski<p>The Budyko framework puts the long-term mean annual evapotranspiration (ET) of a catchment in relation to its maximum possible value determined by the conservation of mass (ET cannot exceed mean annual precipitation) and energy (ET can not exceed mean annual net radiation) in the absence of significant storage contributions. Most catchments plot relatively close to this physical limit, which allowed the development of an empirical equation (often referred to as the Budyko curve) for estimating mean annual evaporation and runoff from observed net radiation and precipitation. Parametric forms of the curve often use a shape parameter, <span class="inline-formula"><i>n</i></span>, that is seen as a catchment characteristic. However, a satisfying explanation for the convergence and self-organization of catchments around such an empirical curve is still lacking. In this study, we explore if vegetation optimality can explain the convergence of catchments along a Budyko curve and in how far can <span class="inline-formula"><i>n</i></span> be seen as a catchment characteristic.</p> <p>The Vegetation Optimality Model (VOM) optimizes vegetation properties and behavior (e.g., rooting depths, vegetation cover, stomatal control) to maximize the difference between the total carbon taken up from the atmosphere and the carbon used for maintenance of plant tissues involved in its uptake, i.e., the long-term net carbon profit (NCP). This optimization is entirely independent of observed ET and hence the VOM does not require calibration for predicting ET. In a first step, the VOM was fully optimized for the observed atmospheric forcing at five flux tower sites along the North Australian Tropical Transect, as well as 36 additional locations near the transect and six Australian catchments. In addition, the VOM was run without vegetation for all sites, meaning that all precipitation was partitioned into soil evaporation and runoff. For comparison, three conceptual hydrological models (TUWmodel, GR4J, and FLEX) were calibrated for the Australian catchments using the observed precipitation and runoff. Subsequently, we emulated step changes in climate by multiplying precipitation (<span class="inline-formula"><i>P</i></span>) by factors ranging between 0.2 and 2 before running the VOM and hydrological models without changing the vegetation properties or model parameters, emulating invariant catchment characteristics under a changed climate. In a last step, the VOM was re-optimized for the different <span class="inline-formula"><i>P</i></span> amounts, allowing vegetation to adapt to the new situation. Eventually, Budyko curves were fit by adapting the parameter <span class="inline-formula"><i>n</i></span> to the model results. This was carried out for both multiple sites simultaneously and for each individual study site, thereby assuming that <span class="inline-formula"><i>n</i></span> is a site-specific characteristic.</p> <p>The optimized VOM runs tracked relatively close to a Budyko curve with a realistic <span class="inline-formula"><i>n</i></span> value and close to observations, whereas the runs without vegetation led to significantly lower evaporative fractions and unrealistically low <span class="inline-formula"><i>n</i></span> values compared with literature. When fitting <span class="inline-formula"><i>n</i></span> to individual catchments, changes in <span class="inline-formula"><i>P</i></span> led to changes in <span class="inline-formula"><i>n</i></span> (increasing <span class="inline-formula"><i>n</i></span> for decreasing <span class="inline-formula"><i>P</i></span>) in all model runs (including the three conceptual models) except if the VOM was re-optimized for each change in <span class="inline-formula"><i>P</i></span>, which brought the value of <span class="inline-formula"><i>n</i></span> back close to its value for the unperturbed <span class="inline-formula"><i>P</i></span> in each catchment. For the re-optimized VOM runs, the variation in <span class="inline-formula"><i>n</i></span> between catchments was greater than within each catchment in response to multiplications of <span class="inline-formula"><i>P</i></span> with a factor 0.2 to 2.</p> <p>These findings suggest that optimality may explain the self-organization of catchments in Budyko space, and that the accompanying parameter <span class="inline-formula"><i>n</i></span> does not remain constant for constant catchment and vegetation conditions as hypothesized in the literature, but in fact emerges through the adaptation of vegetation to climatic conditions in a given hydrological setting. Moreover, the results suggest that <span class="inline-formula"><i>n</i></span> might initially increase in response to suddenly reduced <span class="inline-formula"><i>P</i></span>, and only slowly returns to its original, catchment-specific value, as vegetation re-adjusts to the new climate over decades and centuries. This may constitute a new basis for the evaluation and prediction of catchment responses to climatic shifts.</p>https://hess.copernicus.org/articles/26/6289/2022/hess-26-6289-2022.pdf |
| spellingShingle | R. C. Nijzink S. J. Schymanski Vegetation optimality explains the convergence of catchments on the Budyko curve Hydrology and Earth System Sciences |
| title | Vegetation optimality explains the convergence of catchments on the Budyko curve |
| title_full | Vegetation optimality explains the convergence of catchments on the Budyko curve |
| title_fullStr | Vegetation optimality explains the convergence of catchments on the Budyko curve |
| title_full_unstemmed | Vegetation optimality explains the convergence of catchments on the Budyko curve |
| title_short | Vegetation optimality explains the convergence of catchments on the Budyko curve |
| title_sort | vegetation optimality explains the convergence of catchments on the budyko curve |
| url | https://hess.copernicus.org/articles/26/6289/2022/hess-26-6289-2022.pdf |
| work_keys_str_mv | AT rcnijzink vegetationoptimalityexplainstheconvergenceofcatchmentsonthebudykocurve AT sjschymanski vegetationoptimalityexplainstheconvergenceofcatchmentsonthebudykocurve |