| Soil water content (SWC) is a key variable in the dynamics of a wide variety of soil hydrological processes and in the plant-soil system. For example, SWC mediates the exchange of water and energy between the soil and atmosphere through evaporation and transpiration as well as influences the partitioning of precipitation into runoff, infiltration, and surface water. SWC can be highly variable across space and time, exhibiting variation across physical scales ranging from a few centimeters up to several kilometers and across temporal scales ranging from minutes to years. A better understanding of row-scale SWC heterogeneity is of primary importance for soil water research and agricultural applications such as irrigation programs, precision farming, and hydrological modelling. Such insights can enable the optimization of sampling strategies for measuring SWC and estimating evapotranspiration at the row scale. Furthermore, this research is an essential step toward improving the multi-dimensional modeling of soil water transport within the soil-plant-atmosphere system.This study investigated horizontal distribution characteristics of the SWC at the row scale in corn (Zea mays L.) field in 4 treatments. The treatments included flat planting bed (FPB), flat planting bed with straw mulching on soil surface (FPM), plastic-mulched ridges with bare furrows (RPB), and plastic-mulched ridges with straw-mulched furrows (RPM). The heterogeneity in SWC was assessed from five horizontal measurement points (MPs), at 10-.30-,50-, and 70-cm depths using time domain reflectometry (TDR). The throughfall was measured by a rectangular plexiglass collector at 5 measurement locations (MLs) with the distances of the 0-12 cm (A),12-24 cm (B),24-36 cm (C), 36-48 cm (D) and 48-60 cm (E) from corn row approximately corresponding to plant height (PH) of 50 to 250 cm and the leaf area index (LAI) from 0.4 to 4.0 in 2013 and 2014. The soil surface evaporation was measured by micro-lysimeters (MLS) at 3 MLs with the distances of the 0-12 cm (A),12-24 cm (B) and 24-36 cm (C) from corn row. The field experiments were conducted in 2013 and 2014 in Lishu county. Based on the field experiment we calibrated and validated HYDRUS-2D model to simulate the heterogeneity in soil water fluxes in row scale. The following lists the major results from the research:(1) In FPB and FPM treatments, the heterogeneity in SWC among the five MPs was statistically significant for 19-75% of sampling dates and was most often different between the two treatments at 10-and 30-cm depths. Straw mulch most strongly affected the heterogeneity of SWC at a 10-cm depth, but this effect differed between the two years. With respect to SWC, the FP treatment exhibited a higher range (RA) and lower maximum standard deviation (SD) and coefficient of variation (CV) relative to those of the FPM treatment. In a majority of sampling dates (more than 60%), the most representative SWC estimates were from those MPs positioned at one-quarter into the inter-row gap.(2) In RPB and RPM treatments, the heterogeneity in SWC among the five MPs was statistically significant for 25-81% of sampling dates and was most often different between the two treatments at 10-and 30-cm depths. Straw mulch most strongly affected the heterogeneity of SWC at a 10-cm depth, but this effect differed between the two years. With respect to SWC, the RPB treatment exhibited a higher maximum RA and lower maximum SD and CV than did the RPM treatment. In a majority of sampling dates (more than 20%), the most representative SWC estimates were from those MPs-group comprising B and D MPs for RPB treatment, and the group comprising B and E MPs for RPM treatment.(3) The throughfall (TF) increased with total rainfall amount but decreased with leaf area index (LAI) and plant height values. The lowest TF values were recorded at the MPs of 0-12 cm and 48-60 cm near the corn row, and the highest TF values occurred at the ML of 24-36 cm at the middle of the inter-row. The cumulative TF at the MP of 24-36 cm exceeded that at the MPs of 0-12 cm and 48-60 cm by approximately 88% and 46% in 2013 and 2014, respectively. The coefficient of variation of TF of different MPs was negatively correlated with rainfall amount, but positively correlated with LAI and PH.(4) The heterogeneity in soil surface evaporation amont the different MPs was statistically significant in some sampling dates. The mean daily soil surface evaporation in FPB and RPB treatments were higher than the FPM and RPM treatments. The mean CV of the soil surface evaporation among the different MPs during the sampling dates was almost same among the four threatments in 2013. But the mean CV of the soil surface evaporation among the different MPs during the sampling dates was higher in the RPB and RPM treatments than the FPB and FPM treatments.(5) The water movement was simulated by using HYDRUS-2D model with considering throughfall heterogeneity boundary condition or otherwise in FPB and RPB treatment. The throughfall heterogeneity boundary was obtained by multiple regression analysis based on the throughfall and rainfall amount, LAI and plant height. The parameters were determined with 2013 field data of the FPB and RPB treatments when using HYDRUS-2D model to simulate the soil water movement in corn row scale. In the FPB treatment, the root mean square error (RMSE) between simulated and measured SWC in different positions was ranged from 0.0175-0.0332 cm3 cm-3, the model efficiency (ME) was ranged from 0.70-0.95, the d ranged from 0.94-0.9, and the R2 ranged from 0.90-0.97. In the RPB treatment, the RMSE ranged from 0.0116-0.0387 cm3 cm-3, the ME was ranged from 0.69-0.98, the Pranged from 0.91-1.00, and the R2 ranged from 0.85-0.99. The parameterized model was tested using data of FPB and RPB treatments in 2014. The RMSE between simulated and measured SWC in different positions was ranged from 0.0093-0.0250 cm3 cm-3, the model efficiency (ME) was ranged from 0.37-0.64, the d ranged from 0.67-0.90, and the R2 ranged from 0.37-0.71. In the RPB treatment, the RMSE ranged from 0.0088-0.0243 cm3 cm-3, the ME was ranged from 0.38-0.86, the d ranged from 0.76-0.96, and the R2 ranged from 0.48-0.92.(6) There was difference in soil water fluxes among the different positions in the same soil depths, and the biggest CV in soil water fluxes was in 10 cm and 30 cm depths. The CV of horizontal soil water fluxes among different positons was biggest in 30 cm depth and smallest in 70 cm depth. The CV of the vertical soil water fluxes among different positons was decreased with increasing soil depths.(7) By contrast, the simulated SWC in different positions in corn row by HYDRUS-2D model with considering throughfall heterogeneity boundary condition were closer to the measured SWC than the HYDRUS-2D model without considering throughfall heterogeneity boundary condition. The HYDRUS-2D model simulated with considering throughfall heterogeneity boundary condition can be enhanced the accuracy of simulation in SWC in 10 to 30 cm depths in FPB treatment and 10 to 50 cm depths in RPB treatment, respectively. |
PDF Full Text Request