Simulation Of Water And Heat Transfer In SPAC System For Summer Maize Under Conditions Of Different Water Supply

Quantitative simulation of water and heat transfer in SPAC system under conditions of different water supply can contribute to explain the hydrothermal transfer and transformation processes happening at different interface in soil-plant-atmosphere continum system, to elucidate the water saving mechanism of non-sufficient irrigation method, to effectively resolve the conflicts between crop production and water consumption and ultimately to achieve the efficient use of agricultural water resource. Based on the insufficient irrigation experiment of a summer maize field, this study analyzed the summer maize canopy spectral characteristics and the dynamic variation of stomatal conductance and evapotranspiration. Models to estimate leaf area index and stomatal conductance under different water supply conditions were established to provide dynamic crop parameters driving for the SPAC model. On the basis of STEMMUS model, which fully considered the coupling mechanism of soil water, vapor and thermal transport, the crop evapotranspiration calculation module(direct and indirect ET calculation methods) and root water uptake calculation module(macroscopic and microscopic root water uptake model) were added to develop a water and heat transfer model for SPAC system, to achieve a quantitative simulation of water and heat transfer under different water supply conditions. The main results are outlined as follow:(1) For summer maize, the "green peak", "red ebb" and "red edge" in visible band of canopy hyperspectral reflectance can be used as sensitive bands of leaf area index, the infrared band(1600-1830nm) can be used as a sensitive indicator of canopy water status. Furthermore, we established a model to estimate leaf area index based on three types of hyperspectral vegetation index, namely green vegetation index(NDVI, RVI and EVI), the vegetation index considering soil background correction(MSAVI, TSAVI and PVI) and the water vegetation index(NDWI1240, NDWI1460 and WI). Results indicated that the normalized difference vegetation index NDVI was effective in estimating leaf area index of summer maize in local area, the simulation performance was improved when using the vegetation index considering soil background information, but the improvement was not obvious; using water vegetation index presented a high determination coefficient, which indicated that the simulation performance of leaf area index could be improved significantly by considering canopy water status under different water supply conditions.(2) We used three water stress index(leaf air temperature difference, leaf level crop water stress index, the soil moisture index) to improve Jarvis stomatal conductance model and further analyzed and validated the applicability of the modified Jarvis model under different water supply conditions. The improved Jarvis model based on leaf air temperature difference and leaf level crop water stress index performed obviously better than the improved model based on soil moisture index. In the late growing season of summer maize, the improved Jarvis model based on soil moisture index obviously underestimated stomatal conductance, which was not suitable for the quantitative research of stomatal conductance during this period of summer maize.(3) The diurnal variation of both the evapotranspiration and transpiration rate of summer maize could be characterized as a single peak curve. The effects of different water supply on the transpiration rate mainly reflected on the value and the time of the peak.The response of maize transpiration rate to environmental factors under different water supply conditions were basically the same: the transpiration rate was significantly affected by photosynthetically active radiation(PAR) and air temperature(Ta), less affected by saturation vapor pressure difference(VPD).Under different water supply conditions, the actual evapotranspiration and crop coefficient during summer maize growth period showed a seasonal variation, increased with the development of crop growth and reached the maximum values in the mid-season stage. The effects of different water supply were mainly on the specific values of actual evapotranspiration and crop coefficient.(4) We compared and analyzed the effect of different ET methods(the indirect and direct ET method), different root water uptake models(the macroscopic and microscopic root water uptake model) on the simulation results of the SPAC model in a semi-arid environment,, which provided the basis for the selection of parameterization scheme in the SPAC model under different targets.Simulation results based on macroscopic root water uptake model indicated that:The simulated soil water content values based on two ET methods agreed well with values measured at 20 cm soil depth. However, disagreements increased in deeper soil layers. The simulation of soil temperature performed relatively well for both ET methods. Compared to the observations, the simulation started with good agreement for both ET methods, followed by a significant overestimation after the main irrigation. The disagreement also increased in deeper soil layers. The performance of the two ET methods in estimating hourly, daily and cumulative evapotranspiration differed significantly after irrigation. Using the ETind method resulted in lower values of evapotranspiration. Overall, the ETdir method performed better than the ETind method, the determination coefficient was higher than 0.80.The simulation performance of soil moisture, soil temperature and the evapotranspiration at different time scales based on the microscopic root water uptake model and macroscopic root water uptake model were nearly the same. However, using different ET methods had a remarkable influence on the simulation results of the SPAC model. The ETdir method was preferable for the estimation of maize evapotranspiration in this area.(5) We investigated the applicability of the SPAC model under different water supply conditions. Although some differences existed in the specific values, the SPAC model was able to conduct the quantitative simulation of the variation in soil moisture, evapotranspiration under different water supply conditions. In our study, we found that in the early growing season, reducing the wetting frequency could significantly reduce soil evaporation. Different amount of irrigation exerted a significant effect on the maize transpiration, which indicated that the difference in transpiration is the main reason for the different water consumption under various water supply conditions.