Importance of crop varieties and management practices: evaluation of a process-based model for simulating CO<sub>2</sub> and H<sub>2</sub>O fluxes at five European maize (<i>Zea mays</i> L.) sites

Abstract

Crop varieties and management practices such as planting and harvest dates, irrigation, and fertilization have important effects on the water and carbon fluxes over croplands, and lack or inaccuracy of this information may cause large uncertainties in hydraulic and carbon modeling. Yet the magnitude of uncertainties has not been investigated in detail. This paper provides a comprehensive assessment of the performances of a process-based ecosystem model called ORCHIDEE-STICS (a coupled model between generic ecosystem model ORCHIDEE and the crop growth model STICS), against eddy-covariance observations of CO₂ and H₂O fluxes at five European maize cultivation sites. The results show that ORCHIDEE-STICS has a good potential to simulate energy, water vapor and carbon dioxide fluxes from maize croplands on a daily basis. The model explains 23–75% of the observed daily net ecosystem exchange (NEE) variance at five sites, and 26–79% of the latent heat flux (LE) variance. Similarly, 34–83% of the variance in observed gross primary productivity (GPP) is accounted for by the model. However, only 3–81% of the variance of observed terrestrial ecosystem respiration (TER) is explained. Therefore, simulating TER is shown to be much more difficult than GPP. We conclude that structural deficiencies of the model in the determination of LAI and TER are the main sources of errors in simulating carbon dioxide and water vapor fluxes. A group of sensitivity analyses, by setting different crop variety, nitrogen fertilization, irrigation, and planting date, indicate that any of these factors is able to cause more than 15% change in simulated NEE although the response of these fluxes to management parameters is site-dependent. Varying management practice in the model is shown to affect not only the daily values of NEE and LE, but also the total seasonal cumulative values, and therefore the annual carbon and water budgets. However, LE is found to be less sensitive to management practices than NEE. Multi-site evaluation of the model and sensitivity analysis with respect to management practices performed in this research provide important insights on the model errors for estimating carbon and water vapor fluxes over European croplands

Improving the responses of the Australian community land surface model (CABLE) to seasonal drought

“…This suggests that modifying the model's root water uptake function and (or) incorporating hydraulic redistribution into the model does not impact the model's performance when soil moisture is readily available during the wet season. Most biases in the simulated NEE during the wet season at all three sites result from errors in the simulated ecosystem respiration by CABLE, as also found for other land surface models [ Baker et al , 2008; Li et al , 2011]. During dry seasons, however, the performances of the default CABLE and the other two configurations with either modification of an alternative root water uptake function or a hydraulic redistribution function were much poorer than the simulations including both revised root functions (as in S4).…”

Seasonal and inter-annual variations in carbon fluxes and evapotranspiration over cotton field under drip irrigation with plastic mulch in an arid region of Northwest China

“…Compared to other desert vegetation, cropland is more heavily influenced by anthropogenic activities (e.g. irrigation, fertilization, tillage and other new technologies), thus it is one of the most active com ponents of carbon and water budgets in arid and semiarid regions (Li et al, 2011a).…”

Improved data gap-filling schemes for estimation of net ecosystem exchange in typical East-Asian croplands

“…A large amount of manual biomass sampling could possibly improve the parameterization of daytime NEE, but this sampling method is destructive and thus could change the footprint of the eddy-covariance measurement. Satellite imaging (González-Sanpedro et al, 2008, resolution of 30 m;Jiang et al, 2010, resolution of 1 km), camera retrieving (Migliavacca et al, 2011) or modeling (Li et al, 2011) could be alternatives and expected to reduce these errors and improve the simulation.…”

“…Moreover, croplands are so intensively managed and manipulated by farmers' decisions (e.g. irrigation, different planting and harvesting dates) across both regions and time (Li et al, 2011) that it is difficult to find a universal strategy encompassing the site-specific year-to-year variation. Croplands are usually patchy with a mixture of crop species, which results in mixed NEE information captured by the eddy-covariance technique.…”