Accurate estimation of terrestrial gross primary productivity (
Hybrid modeling of evapotranspiration: inferring stomatal and aerodynamic resistances using combined physics-based and machine learning
Abstract The process of evapotranspiration transfers liquid water from vegetation and soil surfaces to the atmosphere, the so-called latent heat flux ( Q LE ), and modulates the Earth’s energy, water, and carbon cycle. Vegetation controls Q LE by regulating leaf stomata opening (surface resistance r s in the Big Leaf approach) and by altering surface roughness (aerodynamic resistance r a ). Estimating r s and r a across different vegetation types is a key challenge in predicting Q LE . We propose a hybrid approach that combines mechanistic modeling and machine learning for modeling Q LE . The hybrid model combines a feed-forward neural network which estimates the resistances from observations as intermediate variables and a mechanistic model in an end-to-end setting. In the hybrid modeling setup, we make use of the Penman–Monteith equation in conjunction with multi-year flux measurements across different forest and grassland sites from the FLUXNET database. This hybrid model setup is successful in predicting Q LE , however, this approach leads to equifinal solutions in terms of estimated physical parameters. We follow two different strategies to constrain the hybrid model and therefore control for the equifinality that arises when the two resistances are estimated simultaneously. One strategy is to impose an a priori constraint on r a based on mechanistic assumptions (theory-driven strategy), while the other strategy makes use of more observational data and adds a constraint in predicting r a through multi-task learning of both latent and sensible heat flux ( Q H ; data-driven strategy) together. Our results show that all hybrid models predict the target variables with a high degree of success, with R 2 = 0.82–0.89 for grasslands and R 2 = 0.70–0.80 for forest sites at the mean diurnal scale. The predicted r s and r a show strong physical consistency across the two regularized hybrid models, but are physically implausible in the under-constrained hybrid model. The hybrid models are robust in reproducing consistent results for energy fluxes and resistances across different scales (diurnal, seasonal, and interannual), reflecting their ability to learn the physical dependence of the target variables on the meteorological inputs. As a next step, we propose to test these heavily observation-informed parameterizations derived through hybrid modeling as a substitute for ad hoc formulations in Earth system models.
more »
« less
- Award ID(s):
- 2019625
- NSF-PAR ID:
- 10430172
- Date Published:
- Journal Name:
- Environmental Research Letters
- Volume:
- 18
- Issue:
- 3
- ISSN:
- 1748-9326
- Page Range / eLocation ID:
- 034039
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract GPP ) remains a challenge despite its importance in the global carbon cycle. Chlorophyll fluorescence (ChlF) has been recently adopted to understand photosynthesis and its response to the environment, particularly with remote sensing data. However, it remains unclear how ChlF and photosynthesis are linked at different spatial scales across the growing season. We examined seasonal relationships between ChlF and photosynthesis at the leaf, canopy, and ecosystem scales and explored how leaf‐level ChlF was linked with canopy‐scale solar‐induced chlorophyll fluorescence (SIF ) in a temperate deciduous forest at Harvard Forest, Massachusetts,USA . Our results show that ChlF captured the seasonal variations of photosynthesis with significant linear relationships between ChlF and photosynthesis across the growing season over different spatial scales (R 2 = 0.73, 0.77, and 0.86 at leaf, canopy, and satellite scales, respectively;P < 0.0001). We developed a model to estimateGPP from the tower‐based measurement ofSIF and leaf‐level ChlF parameters. The estimation ofGPP from this model agreed well with flux tower observations ofGPP (R 2 = 0.68;P < 0.0001), demonstrating the potential ofSIF for modelingGPP . At the leaf scale, we found that leafF q ’ /F m ’ , the fraction of absorbed photons that are used for photochemistry for a light‐adapted measurement from a pulse amplitude modulation fluorometer, was the best leaf fluorescence parameter to correlate with canopySIF yield (SIF /APAR ,R 2 = 0.79;P < 0.0001). We also found that canopySIF andSIF ‐derivedGPP (GPPSIF ) were strongly correlated to leaf‐level biochemistry and canopy structure, including chlorophyll content (R 2 = 0.65 for canopyGPPSIF and chlorophyll content;P < 0.0001), leaf area index (LAI ) (R 2 = 0.35 for canopyGPPSIF andLAI ;P < 0.0001), and normalized difference vegetation index (NDVI ) (R 2 = 0.36 for canopyGPPSIF andNDVI ;P < 0.0001). Our results suggest that ChlF can be a powerful tool to track photosynthetic rates at leaf, canopy, and ecosystem scales. -
Abstract. The uptake of carbonyl sulfide (COS) by terrestrial plants is linked tophotosynthetic uptake of CO2 as these gases partly share the sameuptake pathway. Applying COS as a photosynthesis tracer in models requires anaccurate representation of biosphere COS fluxes, but these models have notbeen extensively evaluated against field observations of COS fluxes. In thispaper, the COS flux as simulated by the Simple Biosphere Model, version 4(SiB4), is updated with the latest mechanistic insights and evaluated with siteobservations from different biomes: one evergreen needleleaf forest, twodeciduous broadleaf forests, three grasslands, and two crop fields spread overEurope and North America. We improved SiB4 in several ways to improve itsrepresentation of COS. To account for the effect of atmospheric COS molefractions on COS biosphere uptake, we replaced the fixed atmospheric COS molefraction boundary condition originally used in SiB4 with spatially andtemporally varying COS mole fraction fields. Seasonal amplitudes of COS molefractions are ∼50–200 ppt at the investigated sites with aminimum mole fraction in the late growing season. Incorporating seasonalvariability into the model reduces COS uptake rates in the late growingseason, allowing better agreement with observations. We also replaced theempirical soil COS uptake model in SiB4 with a mechanistic model thatrepresents both uptake and production of COS in soils, which improves thematch with observations over agricultural fields and fertilized grasslandsoils. The improved version of SiB4 was capable of simulating the diurnal andseasonal variation in COS fluxes in the boreal, temperate, and Mediterraneanregion. Nonetheless, the daytime vegetation COS flux is underestimated onaverage by 8±27 %, albeit with large variability across sites. On aglobal scale, our model modifications decreased the modeled COS terrestrialbiosphere sink from 922 Gg S yr−1 in the original SiB4 to753 Gg S yr−1 in the updated version. The largest decrease influxes was driven by lower atmospheric COS mole fractions over regions withhigh productivity, which highlights the importance of accounting forvariations in atmospheric COS mole fractions. The change to a different soilmodel, on the other hand, had a relatively small effect on the globalbiosphere COS sink. The secondary role of the modeled soil component in theglobal COS budget supports the use of COS as a global photosynthesis tracer. Amore accurate representation of COS uptake in SiB4 should allow for improvedapplication of atmospheric COS as a tracer of local- to global-scaleterrestrial photosynthesis.more » « less
-
This dataset contains monthly average output files from the iCAM6 simulations used in the manuscript "Enhancing understanding of the hydrological cycle via pairing of process-oriented and isotope ratio tracers," in review at the Journal of Advances in Modeling Earth Systems. A file corresponding to each of the tagged and isotopic variables used in this manuscript is included. Files are at 0.9° latitude x 1.25° longitude, and are in NetCDF format. Data from two simulations are included: 1) a simulation where the atmospheric model was "nudged" to ERA5 wind and surface pressure fields, by adding an additional tendency (see section 3.1 of associated manuscript), and 2) a simulation where the atmospheric state was allowed to freely evolve, using only boundary conditions imposed at the surface and top of atmosphere. Specific information about each of the variables provided is located in the "usage notes" section below. Associated article abstract: The hydrologic cycle couples the Earth's energy and carbon budgets through evaporation, moisture transport, and precipitation. Despite a wealth of observations and models, fundamental limitations remain in our capacity to deduce even the most basic properties of the hydrological cycle, including the spatial pattern of the residence time (RT) of water in the atmosphere and the mean distance traveled from evaporation sources to precipitation sinks. Meanwhile, geochemical tracers such as stable water isotope ratios provide a tool to probe hydrological processes, yet their interpretation remains equivocal despite several decades of use. As a result, there is a need for new mechanistic tools that link variations in water isotope ratios to underlying hydrological processes. Here we present a new suite of “process-oriented tags,” which we use to explicitly trace hydrological processes within the isotopically enabled Community Atmosphere Model, version 6 (iCAM6). Using these tags, we test the hypotheses that precipitation isotope ratios respond to parcel rainout, variations in atmospheric RT, and preserve information regarding meteorological conditions during evaporation. We present results for a historical simulation from 1980 to 2004, forced with winds from the ERA5 reanalysis. We find strong evidence that precipitation isotope ratios record information about atmospheric rainout and meteorological conditions during evaporation, but little evidence that precipitation isotope ratios vary with water vapor RT. These new tracer methods will enable more robust linkages between observations of isotope ratios in the modern hydrologic cycle or proxies of past terrestrial environments and the environmental processes underlying these observations. Details about the simulation setup can be found in section 3 of the associated open-source manuscript, "Enhancing understanding of the hydrological cycle via pairing of process‐oriented and isotope ratio tracers." In brief, we conducted two simulations of the atmosphere from 1980-2004 using the isotope-enabled version of the Community Atmosphere Model 6 (iCAM6) at 0.9x1.25° horizontal resolution, and with 30 vertical hybrid layers spanning from the surface to ~3 hPa. In the first simulation, wind and surface pressure fields were "nudged" toward the ERA5 reanalysis dataset by adding a nudging tendency, preventing the model from diverging from observed/reanalysis wind fields. In the second simulation, no additional nudging tendency was included, and the model was allowed to evolve 'freely' with only boundary conditions provided at the top (e.g., incoming solar radiation) and bottom (e.g., observed sea surface temperatures) of the model. In addition to the isotopic variables, our simulation included a suite of 'process-oriented tracers,' which we describe in section 2 of the manuscript. These variables are meant to track a property of water associated with evaporation, condensation, or atmospheric transport. Metadata are provided about each of the files below; moreover, since the attached files are NetCDF data - this information is also provided with the data files. NetCDF metadata can be accessed using standard tools (e.g., ncdump). Each file has 4 variables: the tagged quantity, and the associated coordinate variables (time, latitude, longitude). The latter three are identical across all files, only the tagged quantity changes. Twelve files are provided for the nudged simulation, and an additional three are provided for the free simulations: Nudged simulation files iCAM6_nudged_1980-2004_mon_RHevap: Mass-weighted mean evaporation source property: RH (%) with respect to surface temperature. iCAM6_nudged_1980-2004_mon_Tevap: Mass-weighted mean evaporation source property: surface temperature in Kelvin iCAM6_nudged_1980-2004_mon_Tcond: Mass-weighted mean condensation property: temperature (K) iCAM6_nudged_1980-2004_mon_columnQ: Total (vertically integrated) precipitable water (kg/m2). Not a tagged quantity, but necessary to calculate depletion times in section 4.3 (e.g., Fig. 11 and 12). iCAM6_nudged_1980-2004_mon_d18O: Precipitation d18O (‰ VSMOW) iCAM6_nudged_1980-2004_mon_d18Oevap_0: Mass-weighted mean evaporation source property - d18O of the evaporative flux (e.g., the 'initial' isotope ratio prior to condensation), (‰ VSMOW) iCAM6_nudged_1980-2004_mon_dxs: Precipitation deuterium excess (‰ VSMOW) - note that precipitation d2H can be calculated from this file and the precipitation d18O as d2H = d-excess - 8*d18O. iCAM6_nudged_1980-2004_mon_dexevap_0: Mass-weighted mean evaporation source property - deuterium excess of the evaporative flux iCAM6_nudged_1980-2004_mon_lnf: Integrated property - ln(f) calculated from the constant-fractionation d18O tracer (see section 3.2). iCAM6_nudged_1980-2004_mon_precip: Total precipitation rate in m/s. Note there is an error in the metadata in this file - it is total precipitation, not just convective precipitation. iCAM6_nudged_1980-2004_mon_residencetime: Mean atmospheric water residence time (in days). iCAM6_nudged_1980-2004_mon_transportdistance: Mean atmospheric water transport distance (in km). Free simulation files iCAM6_free_1980-2004_mon_d18O: Precipitation d18O (‰ VSMOW) iCAM6_free_1980-2004_mon_dxs: Precipitation deuterium excess (‰ VSMOW) - note that precipitation d2H can be calculated from this file and the precipitation d18O as d2H = d-excess - 8*d18O. iCAM6_free_1980-2004_mon_precip: Total precipitation rate in m/s. Note there is an error in the metadata in this file - it is total precipitation, not just convective precipitation.more » « less
-
more » « less
Abstract Carbon, water and energy exchange between the land and atmosphere controls how ecosystems either accelerate or ameliorate the effect of climate change. However, evaluating improvements to processes controlling carbon cycling, water use and energy exchange in global land surface models (LSMs) remains challenging in part because of persistent model errors in estimating leaf area. Here we evaluate the changes in global carbon, water and energy exchange brought about when a LSM prognostic estimates of leaf area are made consistent with estimates from satellites. This approach achieves two aims; first to quantify the effect of ignoring errors in leaf area index (LAI) on land‐atmosphere fluxes and second, to evaluate how closely this LSM replicates fluxes with and without an LAI constraint. We implemented an ensemble Kalman filter with spatiotemporal adaptive inflation to more closely match community land model (CLM5.0) estimates of leaf area to those from the Global Inventory Modeling and Mapping Studies leaf area index (LAI3g) product. We then evaluate the model's estimates of gross primary productivity (GPP) and latent heat flux (LE) against well established global estimates of these fluxes. We find that the model is biased high by 27% relative to the LAI3g product. Moreover, the effect of bias in LAI is substantial for GPP (18%) and LE (6%) and likely to confound efforts to refine processes controlling these fluxes. This data assimilation approach serves as a method to evaluate the efficacy of refinements to flux processes until the processes controlling the dynamics of LAI are better resolved in LSMs.
-
Abstract. Land surface modellers need measurable proxies toconstrain the quantity of carbon dioxide (CO2) assimilated bycontinental plants through photosynthesis, known as gross primary production(GPP). Carbonyl sulfide (COS), which is taken up by leaves through theirstomates and then hydrolysed by photosynthetic enzymes, is a candidate GPPproxy. A former study with the ORCHIDEE land surface model used a fixedratio of COS uptake to CO2 uptake normalised to respective ambientconcentrations for each vegetation type (leaf relative uptake, LRU) tocompute vegetation COS fluxes from GPP. The LRU approach is known to havelimited accuracy since the LRU ratio changes with variables such asphotosynthetically active radiation (PAR): while CO2 uptake slows underlow light, COS uptake is not light limited. However, the LRU approach hasbeen popular for COS–GPP proxy studies because of its ease of applicationand apparent low contribution to uncertainty for regional-scaleapplications. In this study we refined the COS–GPP relationship andimplemented in ORCHIDEE a mechanistic model that describes COS uptake bycontinental vegetation. We compared the simulated COS fluxes againstmeasured hourly COS fluxes at two sites and studied the model behaviour andlinks with environmental drivers. We performed simulations at a global scale,and we estimated the global COS uptake by vegetation to be −756 Gg S yr−1,in the middle range of former studies (−490 to −1335 Gg S yr−1). Basedon monthly mean fluxes simulated by the mechanistic approach in ORCHIDEE, wederived new LRU values for the different vegetation types, ranging between0.92 and 1.72, close to recently published averages for observed values of1.21 for C4 and 1.68 for C3 plants. We transported the COS using the monthlyvegetation COS fluxes derived from both the mechanistic and the LRUapproaches, and we evaluated the simulated COS concentrations at NOAA sites.Although the mechanistic approach was more appropriate when comparing tohigh-temporal-resolution COS flux measurements, both approaches gave similarresults when transporting with monthly COS fluxes and evaluating COSconcentrations at stations. In our study, uncertainties between these twoapproaches are of secondary importance compared to the uncertainties in theCOS global budget, which are currently a limiting factor to the potential ofCOS concentrations to constrain GPP simulated by land surface models on theglobal scale.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1088/1748-9326/acbbe0
