HS10.3

Beech and spruce trees are dominant species in prealpine forests. Thus, plant-specific physiological traits of beech and spruce are key to determine evapotranspiration fluxes from these forests. During dry periods trees adapt to the decreased soil water availability, however these adaptation strategies are not yet well determined by observation data. Adaptations to water limitations are different between species and if not accounted for, may lead to an overestimation of evapotranspiration fluxes. These unknowns add additional uncertainty to the simulation of transpiration patterns with ecohydrological models under water limited conditions. Here we present a comparison of field methods to measure (directly and indirectly) the transpiration process for the purpose of supporting mechanistic ecohydrological modelling. At our mixed beech and spruce forest field site at Waldlabor Zurich we equipped multiple trees with sapflow sensors (hourly measurements) and frequently measured stomatal conductance and leaf water potential (weekly to twice a week) during the 2021 growing season. Along with these plant-physiological measurements, we recorded timeseries of meteorological variables and soil water content and matric potential in different depths (10, 20, 40 and 80cm).

Sapflow measurements suggest that transpiration rates are tightly linked to the magnitude of solar radiation and vapor pressure deficit. Summer transpiration rates were higher in beech trees compared to spruce trees. Most of the early summer of the 2021 growing season was relatively wet, but the months August and September had considerably lower precipitation than the long-term average. This period with low precipitation during August and September led to decreasing soil water content and matric potential, which caused leaf water potentials to decrease accordingly. On the contrary, stomatal conductance remained relatively constant for beech and even increased for spruce, suggesting that under the encountered conditions, stomatal control is not depending directly on leaf water potential. Sapflow rates gradually decreased as the growing season proceeded, but it remains unclear to what degree this decrease was due to phenology, meteorological conditions and/or limited water availability. We compared our measurements to the simulations of an existing mechanistic ecohydrological model (Tethys-Chloris) to test the performance on the observed diurnal dynamics. The comparisons between observed and simulated transpiration rates showed that uncertainties are larger when water availability is limited in the dry periods of August and September.

Our work provides insight into the processes at the soil-plant-atmosphere continuum by the combination of highly resolved measurements and an established mechanistic ecohydrological model. Results highlight how well different measurements of transpiration proxies agree with each other, how suitable they are to assess the actual transpiration rates, and which conditions have larger simulation uncertainties in ecohydrological models and thus need to be better constrained by field observations.

How to cite: Martinetti, S., Floriancic, M., Molnar, P., and Fatichi, S.: Determining transpiration rates from beech and spruce trees with measurements of sapflow, leaf water potential and stomatal conductance, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-12488, https://doi.org/10.5194/egusphere-egu22-12488, 2022.

Estimates of evapotranspiration (ET) which can be derived from in-situ measurements are often difficult to compare because they originate from different research disciplines, were collected at different scales using a range of methods, and they entail method-specific uncertainties.

The BRIDGET toolbox – developed within the Digital Earth project – aims to support the harmonisation and scaling of diverse in-situ ET measurements by providing tools for storage, merging and visualisation of multi-scale and multi-sensor ET data. This requires an appropriate metadata description for the various measurements as well as an assessment of method-specific uncertainties.

BRIDGET is implemented both as a standalone Python package and as part of the existing virtual research environment V-FOR-WaTer. It is organised as a toolbox consisting of several sub-sections which deal with the different in-situ measurement methods, their typical scaling approaches and most relevant analysis functions. A corresponding uncertainty framework is developed separately as a Python package and as a tool in V-FOR-WaTer. Our first focus for BRIDGET is upscaling tree-level sap flow measurements and comparing them to respective transpiration estimates from eddy covariance and lysimeters.

How to cite: Hassler, S. K., Dietrich, P., Kiese, R., Mälicke, M., Mauder, M., Meyer, J., Rebmann, C., Strobl, M., and Zehe, E.: Assessing, scaling and comparing sap flow, eddy covariance and lysimeter measurements in the BRIDGET toolbox, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5178, https://doi.org/10.5194/egusphere-egu22-5178, 2022.

The partitioning of ecosystem evapotranspiration and carbon dioxide fluxes into their plant and ground components is a critical research priority to better understand the water cycle and ecosystem function. Despite advances in different measurement techniques and partitioning models in the last decades, much is still unknown regarding the importance of different components of H₂O and CO₂ fluxes in ecosystems. In this work, we compare three partitioning methods that are based on analysis of conventional high frequency eddy-covariance (EC) data: the flux variance similarity method, the modified relaxed eddy accumulation methods, and the conditional eddy covariance method. First, we test these methods using fields experimental data, comparing them to other reference measurements for the components fluxes (gas chambers and leaf levels measurements). Subsequently, we develop a novel approach for simulating these fluxes in large eddy simulations and apply it to further probe the performance, assumptions, and relative skill of the three methods. The findings allow us to recommend partitioning best practices for their implementation, and to develop methods for the joint analyses of the various approaches.

How to cite: Bou-Zeid, E., Zahn, E., Ghannam, K., Chamecki, M., Katul, G., Thomas, C., and Kustas, W.: Experimental and Numerical Investigation of Flux Partitioning Methods for Water Vapor and Carbon Dioxide, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-10201, https://doi.org/10.5194/egusphere-egu22-10201, 2022.

One of the main issues with obtaining accurate Evapotranspiration (ET) measurements for heterogeneous crops is managing the partition between the contribution of bare soil / cover crop and that of the main crop. Mostly, ET estimates are obtained as an aggregate of the two components, as direct measurements of distinct Evaporation (E) and Transpiration (T) are possible only with high-accuracy and time-costly field lysimeters. Hydrological modelling can provide these kinds of estimates, with the dichotomy between single-source (one energy balance equation for the whole pixel) and two-source (one balance equation each for the vegetated and the non-vegetated pixel fraction) models approaching the problem from different perspectives. In this work, a laboratory lysimeter was employed to obtain disaggregated fluxes from a global ET value and use them to validate the partitioned estimates from a two-source version (FEST-2-EWB) of the single-source FEST-EWB distributed hydrological model, which was also included in the validation as a reference. The lysimeter was sown with grass distributed in three rows, alternated with similar rows of bare soil, with irrigation being provided to the former and not to the latter. Thermal imagery from proximal sensing observations was used to calibrate the models. Two boxes were placed on the lysimeter, one completely vegetated and the other left bare. These boxes were periodically weighted separately from the lysimeter, obtaining accurate measurements of their ET, that were then scaled back to the correspondent areas in the main lysimeter. The model runs, provided similar calibration performances, showed similar global ET values, close to those measured over the lysimeter, but diverged when looking ad transpiration alone. The two-source model offered estimates much closer to those derived from the lysimeter than the single-source model.

How to cite: Paciolla, N., Corbari, C., and Mancini, M.: Lab lysimeter disaggregated ET data for the validation of a two-source model, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5529, https://doi.org/10.5194/egusphere-egu22-5529, 2022.

In light of ongoing global climate change and related increases in extreme hydrological events, it is becoming increasingly important to have a comprehensive knowledge of the ecosystem water cycle to assess ecosystem stability and in agricultural system to ensure sustainable management and food security. Evapotranspiration (ET) plays a crucial role returning up to 90 % of ingoing precipitation back to the atmosphere. In agriculture, further knowledge about plant transpiration (T) and evaporation (E) of different soils could lead to more efficient water use in the future, which will become necessary for agricultural practice in many regions due to climate change related increase in drought events. Here, we wanted to implore impacts of soil types (representing a ful soil erosion gradient) on ecosystem water budgets (ET) and agronomic water use efficiencies (WUEagro).

We conducted a plot experiment with winter rye (September 17, 2020 to June 30, 2021) at the "CarboZALF-D” experimental field which is located in the hilly and dry ground moraine landscape of the Uckermark region in NE Germany. Along an experimental plot (110 m x 16 m) a modern automated gantry crane was built and used for the first time to continuously determine evapotranspiration with two automated chambers. A major advantage of this system is the opportunity to assess management and soil type effects (compared to eddy covariance setups), without corroborating measurement frequency (compared to manual chamber setups).

Three soil types representing the full soil erosion gradient of the hummocky ground moraine landscape (extremely eroded: Calcaric Regosol, strongly eroded: Nudiargic Luvisol, non-eroded: Calcic Luvisol) within each soil type were investigated (randomized block design, 3 replicates per treatment). In addition, we used five different gap-filling methods and compared them in light of their potential to aquire precise water budgets over the entire growth period as well as reproduce short water flux dynamics realistically. The best performance was achieved with methods based on mean-diurnal-variation (MDV) and support vector machine (SVM), including a validation step SVM yielded best predictions of measured ET. Subsequently, we simulated half-hourly ET fluxes and calculated balances of evapotranspiration for the cropping period.

The results show that there are significant differences in evapotranspiration and yield between soil types, resulting in different water use efficiencies (WUEagro). The Calcaric Regosol (extremely eroded) shows a maximum of around 10% lower evapotranspiration and a maximum of around 35% lower water use efficiency (WUE_agro) compared non-eroded soils.  The key period contributing to 50-65 % of overall ET of the entire growth period was from late April until harvest, however differences in the overall ET budget between soil types and manipulation resulted predominantly from small long-term differences between the treatments over the entire growth period.

How to cite: Dubbert, M., Dahlmann, A., Sommert, M., Augustin, J., and Hoffmann, M.: Quantifying evapotranspiration budgets of winter rye using a automated gantry crane – effects of soil type, erosion and management and testing gap filling procedures, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-1188, https://doi.org/10.5194/egusphere-egu22-1188, 2022.

Accurate parametrization and validation of SVAT- or evapotranspiration-models requires robust estimates of transpiration and conductivity on the level of individual leaves. Such estimates are commonly made from measurements with mobile gas exchange systems, which allow precise measurements of leaf transpiration. However, this method has some decisive practical (expensive and labor intensive, both constraining feasible number of replicates) as well as methodological limitations (destruction of the leaf boundary layer). In order to validate the FO3REST model – which estimates the phytotoxic ozone uptake of forest stands – a sensor was required that continuously measures leaf transpiration and conductivity with a high number of replicates. Within the valORTree project, which was carried out from 2019-2021 in the climate chambers of the TUMmesa ecotron facility (Jákli et al. 2021), a novel, low-cost leaf sensor ("TransP") was developed that enables continuous in-situ determination of transpiration and conductivity for the important forest tree species beech (Fagus sylvatica L.) and Norway spruce (Picea abies (L.) H. Karst.) over the entire growing season. The sensor records different temperatures in the leaf/needle environment and was calibrated against the gravimetrically determined transpiration rate (r² = 0.74 for beech; r² = 0.84 for spruce). Measurement inaccuracies can be compensated for by using many of the inexpensive sensors in parallel. The sensor output was validated against measurements using Li-6400 and Li-6800 gas exchange systems (Licor, USA). Differences in the outputs of the two methods could be explained by the fact that the Licor systems measures transpiration based on stomatal conductance, whereas TransP includes the in-situ boundary layer resistance. So far, the sensor has been applied under low-wind conditions in indoor applications and is currently further developed for application in the field.

However, we clearly show that measuring transpiration of beech leaves and spruce needles with the TransP sensor provides robust data. Since TransP operation is minimally invasive and the leaf boundary layer is preserved during measurements, it is assumed that the sensor provides a realistic representation of the in-situ transpiration of individual leaves/needles. In addition, the high temporal resolution of the measurements provides the ability to accurately integrate transpiration over the entire period of the measurement.

Reference

Jákli, B., Meier, R., Gelhardt, U., Bliss, M., Grünhage, L., & Baumgarten, M. (2021). Regionalized dynamic climate series for ecological climate impact research in modern controlled environment facilities. Ecology and evolution, 11(23), 17364-17380.

How to cite: Jákli, B., Goisser, M., Liu, J., and Baumgarten, M.: TransP - a novel sensor for continuous in-situ measurement of transpiration and conductivity at the leaf level, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5302, https://doi.org/10.5194/egusphere-egu22-5302, 2022.

Transpiration (T) makes up the bulk of total evaporation over vegetated land yet remains challenging to predict at landscape-to-global scale.  Model improvements often occur at the expense of model parsimony and an increased dependence on input data that is difficult to acquire at large scale.  T models intended for these scales should ideally be easily scalable using routine meteorological and/or remote sensing data as input.  

Here, we critically evaluate several “big leaf”-type models ranging in their complexity to simulate daily T in a variety of forest biomes.  All these models use input data streams furnished by readily available global reanalysis or satellite-based remote sensing products.   We develop and evaluate a novel moisture stress method based on the Antecedent Precipitation Index (API) serving as proxy for soil moisture supply, motivated by the challenge of acquiring reliable soil moisture and other soil physical property data at large spatial and temporal scales.

We rely on independent estimates of T derived from co-located sap flow and eddy-covariance measurement systems.  The triple collocation technique is employed to quantify error metrics when treating modeled T as a third, independent measurement.

Preliminary results suggests that models that explicitly account for the aerodynamic coupling between canopy surfaces and the atmosphere generally perform better than those that do not, and that the API-based approach to modeling constraints related to soil moisture stress appears as a valid alternative when soil moisture information is unavailable.  

How to cite: Bright, R. M., Miralles, D. G., Poyatos, R., and Eisner, S.: A critical evaluation of simple, flexible, and scalable models of daily transpiration in forested biomes, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7367, https://doi.org/10.5194/egusphere-egu22-7367, 2022.

Evaporation is a key component of the water cycle in the endorheic basins of the Chilean Altiplano. In this study, sub-diurnal to climatological temporal changes of evaporation in a high-altitude saline lake ecosystem in the Atacama Desert are analysed. We analyse the evaporation trends over 70 years (1950-2020) at a high-spatial resolution. The method is based on the downscaling of 30-km hourly resolution ERA5 reanalysis data to 0.1-km spatial resolution data using artificial neural networks. This downscaled data is used in the Penman open water evaporation equation, modified to compensate for the energy balance non-closure and the ice cover formation on the lake during the night. Our evaporation estimates show a consistent agreement with eddy-covariance measurements and reveal that evaporation is controlled by different drivers depending on the time scale. At the sub-diurnal scale, mechanical turbulence is the primary driver. At the seasonal scale, more than 70% of the evaporation variability is explained by the radiative contribution term. At interannual scales, evaporation increased by 2.1 mm per year during the entire study period according to global temperature increases. Last, we find that yearly evaporation depends on the El Niño Southern Oscillation (ENSO), where warm and cool ENSO phases are associated with higher evaporation rates and precipitation rates, respectively. Our results show that warm ENSO phases increase evaporation rates by 15%, whereas cold phases decrease by 2%. This investigation contributes with reliable long-term evaporation estimates over a typical saline lake of an arid region and a replicable methodology for climate change assessment and sustainable water management. 

How to cite: Lobos Roco, F., Hartogensis, O., Suarez, F., Huerta Viso, A., Benedict, I., de la Fuente, A., and Vila-Guereau de Arellano, J.: Multi-scale temporal analysis of actual evaporation on a saline lake in the Atacama Desert, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5021, https://doi.org/10.5194/egusphere-egu22-5021, 2022.

Evapotranspiration (ET) from wetlands is often considered to occur at a potential rate. However, depending on the structural and physiological traits of the dominating plant species, actual ET can deviate substantially from potential ET. Here we present a case study from a restored wetland in north-western Germany which is dominated by moor grass (Molinia caerulea). ET was measured over three years by means of the Bowen Ratio method, leaf transpiration and leaf resistance of moor grass were measured with a porometer, and both green and total leaf area index were measured optically and manually.

Whilst actual and potential ET were practically similar during the period from late summer to the end of winter, they differed significantly from the beginning of spring to early summer and on hot summer days. Two likely reasons for this marked seasonality could be identified. (1) Molinia leaves responded very sensitively to the vapour pressure deficit of the air, independent of the unlimited water supply to its roots. (2) A thick mat of dead leaves covered the surface in spring before and while the new leaves emerged and acted as an efficient protection cover against evaporation.

The SVAT model ‘AMBAV’ was developed by the German Meteorological Service and is operationally used in agrometeorological applications. Based on the Bowen Ratio and in-situ plant physiological data, it was newly parameterised for the investigated type of wetland. If run with weather data from a nearby station, AMBAV could verify the observed seasonal pattern of actual ET from the moor grass dominated wetland. The results demonstrate that the present vegetation reduces wetland ET and thus contributes to the maintenance of a high water table in the restored wetland.

How to cite: Herbst, M., Matuschek, D., and Falge, E.: Assessing the influence of the vegetation on the evapotranspiration from a wetland – a case study from northwest Germany based on in-situ measurements at different scales, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-7002, https://doi.org/10.5194/egusphere-egu22-7002, 2022.