Accurate estimation of root-zone soil moisture (SM) is of great significance for accurate irrigation management. This study was purposed to identify planted-by-planted mapping of root-zone SM on three critical fruit growth periods based on UAV multispectral images using three machine learning (ML) algorithms in a kiwifruit orchard in Shaanxi, China. Several spectral variables were selected based on variable importance (VIP) rankings, including reflectance Ri at wavelengths 560, 668, 740, and 842 nm. Results indicated that the VIP method effectively reduced 42 vegetation indexes (VIs) to less than 7 with an evaluation accuracy of root-zone SM models. Compared with deep root-zone SM models (SM40 and SM60), shallow root-zone SM models (SM10, SM20, and SM30) have better performance (R² from 0.65 to 0.82, RRMSE from 0.02 to 0.03, MAE from 0.20 to 0.54) in the three fruit growth stages. Among three ML algorithms, random forest models were recommended for simulating kiwi root-zone SM during the critical fruit growth period. Overall, the proposed planted-by-planted root-zone SM estimation approach can be considered a great tool to upgrade the toolbox of the growers in site-specific field management for the high spatiotemporal resolution of SM maps.

The temporal and spatial distributions of regional irrigation water productivity (RIWP) are crucial for making decisions related to agriculture, especially in arid irrigated areas with complex cropping patterns. Thus, in this study, we developed a new RIWP model for an irrigated agricultural area with complex cropping patterns. The model couples the irrigation- and drainage-driven soil water and salinity dynamics and shallow groundwater movement in order to quantify the temporal and spatial distributions of the target hydrological and biophysical variables. We divided the study area into 1 km × 1 km hydrological response units (HRUs). In each HRU, we considered four land use types: sunflower fields, wheat fields, maize fields, and uncultivated lands (bare soil). We coupled the regional soil hydrological processes and groundwater flow by taking a weighted average of the water exchange between unsaturated soil and groundwater under different land use types. The RIWP model was calibrated and validated using 8 years of hydrological variables obtained from regional observation sites in a typical arid irrigation area in North China, the Hetao Irrigation District. The model simulated soil moisture and salinity reasonably well as well as groundwater table depths and salinity. However, overestimations of groundwater discharge were detected in both the calibration and validation due to the assumption of well-operated drainage ditch conditions; regional evapotranspiration (ET) was reasonably estimated, whereas ET in the uncultivated area was slightly underestimated in the RIWP model. A sensitivity analysis indicated that the soil evaporation coefficient and the specific yield were the key parameters for the RIWP simulation. The results showed that the RIWP decreased from maize to sunflower to wheat from 2006 to 2013. It was also found that the maximum RIWP was reached when the groundwater table depth was between 2 and 4 m, regardless of the irrigation water depth applied. This implies the importance of groundwater table control on the RIWP. Overall, our distributed RIWP model can effectively simulate the temporal and spatial distribution of the RIWP and provide critical water allocation suggestions for decision-makers.