Evaluating the E 3 SM Land Model at a temperate forest site using flux and soil water measurements

Accurate simulations of soil respiration and carbon dioxide (CO2) efflux are critical to project global biogeochemical cycles and the magnitude of carbon (C) feedbacks to climate change in Earth system models (ESMs). 15 Currently, soil respiration is not represented well in ESMs, and few studies have attempted to address this deficiency. In this study, we evaluated the simulation of soil respiration in the Energy Exascale Earth System Model (E3SM) using long-term observations from the Missouri Ozark AmeriFlux (MOFLUX) forest site in the central U.S. Simulations using the default model parameters significantly underestimated annual soil respiration and gross primary production, while underestimating soil water potential during growing seasons and overestimating it during non-growing seasons. A site-specific soil water 20 retention curve significantly improved modelled soil water potential, gross primary production and soil respiration. However, the model continued to underestimate soil respiration during peak growing seasons, and overestimate soil respiration during non-peak growing seasons. One potential reason may be that the current model does not adequately represent the seasonal cycle of microbial organisms and soil macroinvertebrates, which have high biomass and activity during peak growing seasons and tend to be dormant during non-growing seasons. Our results confirm that modelling soil respiration can be 25 significantly improved by better model representations of the soil water retention curve.


Introduction
Globally, soils store over twice as much carbon (C) as the atmosphere (Chapin III et al., 2011).Soil respiration (SR) is the second largest C flux between terrestrial ecosystems and the atmosphere (Luo and Zhou, 2006).An accurate simulation of SR is critical for projecting terrestrial C status, and therefore climate change, in Earth system models (ESMs) (IPCC, 2013).efflux to the atmosphere still have a high degree of uncertainty (Friedlingstein et al., 2006;Jones et al., 2013;Todd-Brown et al., 2013;Todd-Brown et al., 2014;Tian et al., 2015), calling for comprehensive assessments of model performance against observational data.
To assess the performance of ESMs, different types of data can be used.For example, using atmospheric CO2 measurements, eddy covariance measurements and remote sensing images, Randerson et al. (2009) found that two ESMs underestimated net C uptake during the growing season in temperate and boreal forest ecosystems, primarily due to the delays in the timing of maximum leaf area in the models.By comparing remote sensing estimations from the Moderate Resolution Imaging Spectroradiometer and flux tower datasets, Xia et al. (2017) found that better representations of processes controlling monthly maximum gross primary productivity (GPP) and vegetation C use efficiency (CUE) improved the ability of models to predict the C cycle in permafrost regions.
Despite the significance of large global SR fluxes, SR has rarely been evaluated in ESMs using long-term observations.Among the factors that influence SR, soil water potential (SWP) provides a unified measure of the energy state of soil water that limits the growth and respiration of plants and microbes.Unlike soil temperature (ST) or soil volumetric water content (VWC), however, SWP is difficult to directly monitor in the field.Accurate estimation of SWP largely relies on the soil water retention curve (i.e., the relationship between VWC and SWP), which is highly specific to soil properties (Childs, 1940;Clapp and Hornberger, 1978;Cosby et al., 1984;Tuller and Or, 2004;Moyano et al., 2013).Site-level data have been used to evaluate model representations of other processes, such as phenology, net primary production (NPP), transpiration, leaf area index (LAI), water use efficiency, and nitrogen use efficiency (Richardson et al., 2012;De Kauwe et al., 2013;Walker et al., 2014;Zaehle et al., 2014;Mao et al., 2016;Duarte et al., 2017;Montané et al., 2017).In Powell et al. (2013), the only aspect concerning SR was the sensitivity of SR to VWC in an Amazon forest, but the study resulted in no improvements to simulated SR.Here, we focus on improving simulations by using site-specific measurements to assess multiple factors influencing SR.
We will evaluate the simulation of SR step by step.We assessed underlying mechanisms in the Energy Exascale Earth System Model (E3SM) by using intensive observations at the Missouri Ozark AmeriFlux (MOFLUX) forest site in the central U.S. We first evaluated the effects of two abiotic factors, ST and SWP, on the simulation of SR.Then we evaluated the effects of biotic factors, such as GPP, LAI and Q10 of heterotrophic respiration, on the simulation of surface CO2 efflux to the atmosphere.

Study site and measurements
The MOFLUX site is located in the University of Missouri's Thomas H. Baskett Wildlife Research and Education Area (latitude 38º44'39"N, longitude 92º12'W).The mean annual precipitation is 1083 mm, while minimum and maximum monthly mean temperatures are −1.3 ºC (January) and 25.2 ºC (July), respectively.The site is a temperate, upland oak- hickory forest, with major tree species consisting of white oak (Quercus alba L.), black oak (Q.velutina Lam.), shagbark hickory (Carya ovata (Mill.)K. Koch), sugar maple (Acer saccharum Marsh.), and eastern red cedar (Juniperus virginiana L.) (Gu et al., 2016;Wood et al., 2017).The dominant soils are the Weller silt loam and the Clinkenbeard very flaggy clay loam (Young et al., 2001).
Ecosystem C, water and energy fluxes, SR, LAI and supporting meteorological measurements were initiated in June 2004 (Gu et al., 2016).Soil respiration was measured within the ecosystem flux footprint using non-flow through non-steady state auto-chambers.From 2004 through 2013, SR was measured using eight automated, custom-built, chambers (ED system; Edwards and Riggs, 2003;Gu et al., 2008) coupled with an infrared gas analyzer (LI-820 Li-Cor Inc., Lincoln, Nebraska).In 2013, this system was replaced with 16 auto-chambers operated using the closed-path system (model LI-8100; Li-Cor Inc., Lincoln, Nebraska).The two systems (ED and Li-8100) were operated side-by-side for several weeks in 2010 and found to produce comparable responses (Paul Hanson, personal communication).Half-hourly SR time series were generated to coincide with the ecosystem flux data set by averaging those chambers sampled in the corresponding averaging period.Net ecosystem CO2 exchange (NEE) was measured on a 32-m walk-up scaffold tower (Gu et al., 2016).A soil temperature profile sensor (model STP01, HuksefluxUSA, Inc., Center Moriches, NY) measured at 5 depths down to 0.5 m.
Soil VWC was measured using water content reflectometers (model CS616, Campbell Scientific Inc., Logan UT) installed beneath each soil chamber.All the data were recorded at half-hourly intervals, which were integrated over time to obtain daily and annual fluxes.

Ecosystem C flux partitioning
Flux-tower GPP was estimated from measured NEE.To reduce biases resulting from individual methods, three NEEpartitioning approaches were employed.The average and variation of the three methods were used to evaluate the modelsimulated GPP.In the first two methods, ecosystem respiration (ER) was estimated from nighttime NEE and extrapolated to daytime, and daytime GPP was calculated from NEE and the extrapolated ER (Reichstein et al., 2005).The only difference between the two methods was whether to exclude night-time data under non-turbulent conditions.In the third method, GPP was estimated by fitting the light-response curve between NEE and radiation (Lasslop et al., 2010).All the partitioning calculations were conducted using the R package REddyProc (Reichstein et al., 2017).

Model description
The E3SM includes coupled carbon and nitrogen cycles, with 10 belowground layers in the biogeochemical module.The land model of E3SM was developed from Community Land Model (CLM 4.5) (Oleson et al., 2013).We implemented offline E3SM Land Model (ELM) simulations with the CLM-CN soil decomposition module (Thornton and Rosenbloom, 2005;Bonan et al., 2013).In the model, SR was calculated by different CO2 emission components (Oleson et al., 2013) Eq. ( 7) where RA and RH are belowground autotrophic and heterotrophic respiration, respectively.RA is the sum of root maintenance (RM) and growth respiration (RG).Rlivecroot and Rfroot are maintenance respiration of live course root and fine root.
[N]livecroot and [N]froot are nitrogen content of live coarse and fine roots.Rbase is the base maintenance respiration at 20 ºC.Rq10 which equals 2, is the temperature sensitivity of maintenance respiration.T 2m is the air temperature at 2 m.C new_root is the new root growth C. RH is the sum of heterotrophic respiration of four SOC pools with different turnover rates (Oleson et al., 2013) in the 10 soil layers.The parameters ki and rfi are the turnover rate and respiration fraction of the i th pool.T, W,O,D,N are environmental modifiers of soil temperature, soil water content, oxygen, depth and nitrogen for each layer, respectively.A detailed description of the environmental modification can be found in Oleson et al. (2013).Briefly, the temperature and water modifiers were: Eq. ( 8) where Q10 is the temperature sensitivity (the default value is 1.5), Tref is the reference temperature (25 ºC).m is the matric water potential, min is the lower limit for matric potential, and s is the matric water potential under saturated conditions.
The E3SM is a grid-based ESM.To assess it using site-level observations, we used a point-run framework which allows the model to simulate individual sites (Mao et al., 2016).With site-specific forcing, a 200-year accelerated decomposition spinup was performed, followed by a 200-year normal spin-up, before the transient simulation was performed from 1850 to 2013.
The vegetation was set as 100% temperate deciduous forest.

Soil water retention curve
Soil water potential values for the Weller soils (https://soilseries.sc.egov.usda.gov/OSD_Docs/W/WELLER.html) were estimated from observed VWC and soil water retention curves that were developed for the site.To derive the soil water retention curves, soil samples were collected in the area of the flux tower base at two depths: 0 to 30 cm and below 30 cm.Samples were evaluated periodically for soil water potential using a dewpoint potentiometer (Decagon Devices, Model WP4C) as they dried over time (Hanson et al., 2003).
In the Brooks & Corey model, the SWP-VWC relationship was expressed as Eq. ( 14) where   and   are the residual and saturated water contents, respectively,  and Ψ  are measured VWC and matric potential (MPa), Ψ  is a parameter related to the soil matric potential at air entry, and  is related to the soil pore size distribution (Brooks and Corey, 1964).
In the Fredlund & Xing model, the SWP-VWC relationship was described as where a, n and m are parameters determining the shape of the soil water characteristic curve (Fredlund and Xing, 1994).
In the Hanson model (Hanson et al., 2003), soil matric potential was modeled by a double exponential function: where a, b, c and d are fitted parameters.
In the van Genuchten model, the SWP-VWC relationship was described as Eq. ( 17) where  (MPa -1 ) and n are parameters that determine the shape of the soil-water curve (van Genuchten, 1980).In addition to the default SWP-VWC relationship in the E3SM, all the five empirical models were parameterized using non-linear fitting against measured VWC and SWP data from the study site.The root-mean-square error (RMSE) was used to select the best model representing the SWP-VWC relationship, where smaller RMSE implies a better fit to observational data.The best-fit model was used in two ways.First, it was used to calculate the "observed" SWP from monitored VWC in the field.Second, it was implemented in the E3SM to replace the default SWP model in order to improve the SWP simulation.

Evaluation of SR in the model
The evaluation of SR was conducted step by step.We first compared observations with the model default output of SR and related factors, including ST, SWP, GPP, and LAI.Thereafter, we attempted to improve the simulation of these factors in order to improve the overall SR simulation by (i) implementing the best-fit SWP-VWC relationship, and (ii) modifying model parameters related to GPP, LAI and SR.GPP-related parameters included the specific leaf area (SLA) at the top of canopy and the fraction of leaf nitrogen in the RuBisCO enzyme.LAI-related parameters included the number of days to complete leaf fall during the end of growing season, the critical day length for senescence (i.e., the length of the day when leaves start to senesce), and a parameter  that was used to produce a linearly-increasing rate of litterfall.In addition, the Q10 of heterotrophic respiration was also modified.Because the parameter modification was dependent on the evaluation steps, how the parameters were modified is presented in the Results section.

Results
For the upper 30 cm of soil, E3SM simulations using the default Clapp and Hornberger model tended to underestimate the SWP when VWC was less than 15% (Fig. 1a), while SWP rapidly approached zero when VWC was greater than 25% (Fig. 1a).For soil below 30 cm, the E3SM showed a consistent overestimation of SWP when VWC exceeded 15% (Fig. 1b).The default E3SM showed relatively high RMSE for both soil layers, indicating that the SWP-VWC relationship was not well simulated in the E3SM (Table 1).Although the Clapp & Hornberger model performed better by using parameters from nonlinear fitting, its performance was not as good as the Hanson and the van Genuchten models (Table 1, Fig. 1).The E3SM default run significantly underestimated both annual SR and GPP (Fig. 2).In addition, the simulated SR had smaller interannual variability compared to the observations.The model was not able to simulate the steep drop of SR or GPP during the extreme drought in 2012.The simulations of ST and SWP were isolated to analyse their contributions to model performance.Whereas the model simulated ST well at 10 cm depth (Fig. 3a), it tended to underestimate SWP when water is limiting and to overestimate SWP otherwise (Fig. 3b).Implementing the data-constrained Hanson model significantly improved the simulation of SWP, showing a greater R 2 and a much smaller RMSE than that of the default run (Fig. 3b).After improving the simulation of SWP, the model better matched the observed annual SR and GPP (Fig. 2).The mean annual simulations of SR and GPP fell into the 95% confidence interval of observations (red bar in the inserted plot in Fig. 2).In addition, the changes in annual SR and GPP (i.e., the differences between before and after the improved SWP simulation) showed a linear relationship (Fig. 4).Despite the improved simulation of SR, the model still underestimated SR and GPP during peak growing seasons when SR and GPP were high, but overestimated them during non-peak growing seasons (Figs. 5, Fig. S1).In other words, though 5 the improved simulation of SWP increased SR and GPP during peak growing seasons, the model still showed systematic errors (Fig. 5).We attempted to improve the seasonal simulations of SR, GPP and LAI by modifying several related parameters (Table 2).Using measurements of NEE, maximum LAI and latent heat fluxes from the MOFLUX site, Lu et al.
(unpublished) calibrated a polynomial surrogate model of CLM4.5.Based on their results, we modified two parameters, i.e., the SLA at the canopy top from 0.03 to 0.01, and the fraction of leaf nitrogen in the RuBisCO enzyme from 0.1007 to 0.12. 10  Comparing the simulated LAI with the observations (Fig. 6), we found that the parameter ndays_off (number of days to complete leaf offset) in the E3SM was too short (default value = 15 days) for the MOFLUX site.Thus, we reset the value of ndays_off to 45 days.We also modified the values of two additional parameters, i.e., crit_dayl and correspondingly (Table 2) Parameter crit_dayl (the critical day length for senescence, units: second) triggers the leaf falling during the end of the growing season.Parameter () is used to produce a linearly-increasing litter-fall rate.The adjustment of the aforementioned five parameters (Table 2) significantly reduced the LAI to within a more reasonable range, showing an increased R 2 from 0.05 to 0.91 and a reduced RMSE from 1.15 to 0.34 (Fig. 6a).However, the simulation of GPP was not improved, though the simulation during peak growing season was increased (Fig. 6b).The model was still not able to reach the high values of SR observed during peak growing seasons, even when the modeled GPP exceeded the observations (Fig. 6b, c).In addition, the modeled SR during non-peak growing seasons was not improved (Fig. 6c).

Effect of SWP on annual SR
Constraining the SWP-VWC relationship with site-specific data and using the Hanson model instead of the E3SM default model (Fig. 1) significantly improved the model representation of SWP (Fig. 3).The modeled annual SR in the E3SM was also substantially improved (Fig. 2a).Although SWP can directly affect SR, the improved annual SR simulation was likely due to the indirect influence of SWP on GPP.The land model of E3SM underestimated GPP in this study (Fig. 2b).
Similarly, a recent study showed that the CLM4.5 significantly underestimated GPP at a coniferous forest in northeastern United States as well (Duarte et al., 2017).GPP affects the substrate supply for SR, as evidenced by the close relationship between changes in SR and GPP (Fig. 4), which is consistent with experimental evidence showing GPP can directly affect the magnitude of SR (Craine et al., 1999;Högberg et al., 2001;Wan and Luo, 2003;Verburg et al., 2004;Gu et al., 2008).
The simulation of SWP in the default E3SM was poor compared with that of ST (Fig. 2), which may be a common issue in ESMs (Powell et al., 2013;Reyes et al., 2017).Powell et al. (2013) showed that differential sensitivity of SR to VWC in several ESMs using observations in two Amazon forests.Our analyses in this study indicate that improving the modeled SWP can significantly improve SR simulations.Thus, we argue that the SWP simulation in ESMs should be calibrated carefully with observations, and/or by using different model representations of the SWP-VMC relationship.
In this study, we derived better SWP-VWC relationship by using non-linear fitting, primarily because of the availability of soil moisture retention curve data.It is an efficient method when site-level data is available, but it is not realistic to calibrate the water retention curve for every site.The SWP-VWC relationship is dependent on soil texture (Clapp and Hornberger, 1978;Cosby et al., 1984;Tuller and Or, 2004), so building relationships between model parameters and soil texture may allow efficient extrapolations of site-level measurements to regional and global scales.
Parameters in the default Clapp & Hornberger model used in the E3SM were derived from synthesizing data across soil textural classes (Clapp and Hornberger, 1978;Cosby et al., 1984;Lawrence and Slater, 2008).The data were derived from over 1,000 soil samples from 11 USDA soil textural classes (Holtan et al., 1968;Rawls et al., 1976).The dependence of model parameters on soil texture were derived from a regression of these 11 data points, i.e., the mean parameter values of 11 soil textural classes against the sand or clay fractions (Cosby et al., 1984).Because no actual sand or clay content of soil samples was reported in the original databases (i.e., only the soil textural classes were reported), the sand and clay fractions used for the regression were obtained from midpoint values of each textural class (Clapp and Hornberger, 1978;Cosby et al., 1984).One potential issue is that soil samples in the same textural classes can have different sand and clay contents and In addition, different empirical models have been developed to describe the SWP-VWC relationship (Brooks and Corey, 1964;Clapp and Hornberger, 1978;van Genuchten, 1980;Fredlund and Xing, 1994;Hanson et al., 2003).These models could be evaluated with an updated SWP-VWC database, and the selected best-fit model(s) could be used to calculate SWP in the field from continuously monitored VWC (e.g., from AmeriFlux network) on different spatial and temporal scales.A new field SWP-VWC database at different scales could be used as a benchmark to improve simulations of soil water and biogeochemical processes in ESMs.

Improving the representation of seasonal variability of SR
Although the simulation of the SWP improved the representation of both annual SR and GPP, the E3SM continued to underestimate SR during peak growing seasons and overestimate SR during other seasons (Figs. 5, 6c, Fig. S1).Since it has been demonstrated that GPP can directly influence the magnitude of SR (Craine et al., 1999;Högberg et al., 2001;Wan and Luo, 2003;Verburg et al., 2004), the underestimated SR during peak growing seasons may be due to an underestimation of GPP.However, our analyses showed that the modeled SR was not able to reach the observed peak in most years even when the modeled GPP exceeded the observation (Fig. 6).In addition, the parameter modification increased GPP during both peak and non-peak growing seasons, resulting in an even greater overestimation of SR during non-peak growing seasons (Fig. 6).
These results suggest that simply increasing GPP may not be adequate to increase the seasonal variability of the simulated SR.Specifically, heterotrophic respiration contributed a majority proportion (i.e., over 85%) to total SR during non-growing seasons (Fig. 7), suggesting that the overestimation of SR during these seasons was primarily due to the biased heterotrophic respiration simulation.
A potential reason for the biased heterotrophic respiration simulation may be related to the temperature sensitivity (Q10).
Theoretically, a higher Q10 can result in greater seasonal variability of SR (Fig. S2).Compared to relatively small Q10 values, a larger Q10 can lead to lower heterotrophic respiration when temperature is below the reference temperature, and greater heterotrophic respiration when temperature is above the reference (Fig. S2).In the E3SM, the reference temperature is 25 ºC and the Q10 of heterotrophic respiration is 1.5 (Oleson et al., 2013).A previous study derived a much greater Q10 value (i.e., 2.83) when the parameters were calibrated with data from another temperate forest (Mao et al., 2016).We hypothesized that the Q10 value of 1.5 may be too small for the MOFLUX site.We arbitrarily increased Q10 from 1.5 to 2.5, but there were minimal effects on the SR simulation (Fig. S3).This indicates that modifying the temperature sensitivity of heterotrophic respiration may not improve the modeled representation of seasonality of SR in the E3SM.Another potential reason for the biased heterotrophic respiration simulation may be that the seasonality of microbial organisms was not adequately represented in the model.Like most ESMs, the E3SM represents soil C dynamics using linear differential equations and assumes that SR is a substrate-limited process in the model.However, producers of CO2 in soils, microbial organisms, have a significant seasonal cycle (Lennon and Jones, 2011).These organisms usually have very high biomass and activity during growing season peaks with favorable conditions of temperature, moisture and substrate supply, and tend to be dormant under stressful conditions (Lennon and Jones, 2011;Stolpovsky et al., 2011;Wang et al., 2014;Wang et al., 2015).The seasonality of microbial biomass and activity, in addition to that of GPP and ST, may contribute to the seasonal variability of SR.
Additionally, lacking representation of macroinvertebrate and other forest floor and soil fauna in the E3SM may be another reason.There is a high density of earthworms at the MOFLUX site (Wenk et al., 2016).Earthworms can shred and redistribute soil C and change soil aggregation structure, which may alter soil C dynamics and CO2 efflux to the atmosphere (Verhoef and Brussaard, 1990;Brussaard et al., 2007;Coleman, 2008).Like microbial organisms, earthworms usually have a significant seasonal cycle, showing high biomass and high activity during peak growing seasons and tending to be dormant during non-growing seasons (Wenk et al., 2016).However, a recent review suggests that current experimental evidence and conceptual understanding remains insufficient to support the development of explicit representation of fauna in ESMs (Grandy et al., 2016).Therefore, data collection focused on seasonal variations in fauna and microbial biomass and activity might enable further improvements in the representation of seasonal variation in SR.

Conclusions
In this study, we used temporally extensive and spatially distributed site observations of SR to assess the capabilities of E3SM.These results indicated that an improved representation of SWP within the model provided better simulations of annual SR.This underscores the need to calibrate SWP in ESMs for more accurate projections of coupled climate and biogeochemical cycles.Notwithstanding this improvement, however, the E3SM still underestimated the magnitude of SR during the peak growing season and overestimated SR during the non-peak growing season.Additional adjustments of driving variables for SR, including GPP, LAI and the Q10 of heterotrophic respiration did not improve the model prediction.
It may be that inadequate model representation of the seasonality of fauna and microbial organisms could be explored as means to achieve better fit.Future incorporation of explicit microbial processes with relevant data collection activities ma y therefore enable improved model simulations.

Figure 1 :
Figure 1: Observed (black dots) and simulated relationship between soil water potential (SWP) and volumetric water content (VWC) by the different models at two soil layers: (a) 0 to 30 cm and (b) below 30 cm.

Figure 2 :
Figure 2: Annual soil respiration (SR) and gross primary production (GPP).Blue and red lines are model outputs before (MODdefault) and after (MODswp) soil water potential improvement, respectively.Black lines and grey area are the observed (OBS) mean and 95% confidence interval (CI), which were calculated from 8 field replications for SR, and from three different net ecosystem exchange partitioning methods for GPP.The inserted bar plots are mean annual average ± 95% CI across 2005-2011 for MODdefault (blue), MODswp 5

Figure 4 :
Figure 4: Relationship between changes in simulated annual soil respiration (SR) and gross primary production (GPP) induced by improvement of soil water potential in the model.

Figure 5 :
Figure 5: Observed vs. modeled soil respiration (SR) and gross primary production (GPP).Blue and red dots are model outputs before (MODdefault) and after (MODswp) soil water potential improvement, respectively.The black line is 1:1 line.

Figure 6
Figure 6 Daily leaf area index (LAI), gross primary production (GPP) and soil respiration (SR).OBS: observation; MODdefault: model output before soil water potential improvement; MODswp_param: model output after soil water potential improvement and parameter modification.a, the blue (MODdefault) and red (MODswp_param) arrows are used to illustrate the earlier start of litterfall in MODswp_param.
ESMs.For example, using a reduced-complexity model, Todd-Brown et al. (2013) demonstrated that the spatial variation in soil C in most ESMs is primarily dependent on C input (i.e., NPP) and ST, showing R 2 values between 0.62 and 0.93 for 9 of 11 ESMs.However, the same reduced-complexity model, driven by observed NPP and ST, can only explain 10% of the variation in the Harmonized World Soil Database observational database (Todd-Brown et al., 2013).These previous results indicate that other important factors affecting soil C dynamics, in addition to NPP and ST, are inadequately simulated in Geosci.Model Dev.Discuss., https://doi.org/10.5194/gmd-2018-34Manuscript under review for journal Geosci.Model Dev. Discussion started: 5 March 2018 c Author(s) 2018.CC BY 4.0 License.SWP-VWC relationships, which may not be fully represented when they are grouped together.Consequently, the parameterized Clapp & Hornberger model may still perform poorly for specific sites, e.g., the MOFLUX site.A re-analysis of an updated SWP-VWC database, with actual sand and clay content measurements, may enable improved relationships between model parameters and soil texture in the water retention model.

Table 1 .
The Hanson model was the best-fit model for the MOFLUX site, showing the smallest RMSE for both soil layers (Table1, Fig.1), and was therefore implemented in E3SM to calculate SWP from measured VWC.Root-mean-square-error (RMSE) of different models in simulating the SWP-VWC relationship for the soil in the MOFLUX site at two depths: 0 to 30 cm and below 30 cm.

Table 2 .
Modified parameters to better simulate gross primary production (GPP) and leaf area index (LAI) at the MOFLUX site in the E3SM.