Discussion

Impact of infiltration from precipitation or flood irrigation on soil water

The soil water is significantly influenced by the precipitation, evapotranspiration, and infiltration, showing that the SWPs and SWCs increase during the irrigation /precipitation period and decrease during the non-irrigation period (Fig. 3 and 4). The SWCs and SWPs before and after the 59.2 mm large rain event during the period 25-27 June 2019, the 68mm irrigation on 3 August 2019, the 137mm irrigation on 6 July 2020 and the 122m irrigation on 1 August 2020, were used to evaluate the process of precipitation /irrigation infiltration and soil water movement (Fig. 4).
At the surface layer (0-50 cm), it was generally found that the SWCs increased from the range of 0.18-0.33 to the range of 0.25-0.37 after the large rain. The average SWP at 0-50cm increased from -120 to -56 cm and the values below a depth of 50 cm were relatively stable. Similarly, soil water isotopic compositions were depleted in δD and δ18O in soil layers above 50 cm in response to the large rainfall event. The average value δD at 0-50cm ranged from −37.8‰ to −53.4‰ and δ18O from −6.17‰ to −7.64‰, respectively. Main changes in the SWCs, SWPs and isotopic compositions at 0-50cm depth during precipitation and non-precipitation periods was greater than those at other depths. Therefore, it can be identified that the soil at 0-50 depth with bare ground surface is mainly influenced by precipitation infiltration and evaporation to a greater extent compared with below. These results correspond to the findings of Wang et al., (2012), who found that the depth mainly affected precipitation infiltration and evaporation was 0-50cm in Yuncheng, Shanxi Province.
The SWCs exhibited a sharp increase when irrigation began and then decreased rapidly as it was turned off. The increase in SWCs occurred layer by layer from the upper horizons, suggesting piston flow was the dominant mode of irrigation-driven soil water flow. The soil water δD and δ18O in 2019 and 2020 had almost the same range as irrigation water, which indicated rapid infiltration of soil water across the entire profile to irrigation (Fig. 4). After 6 days of irrigation, the SWCs at 20 and 30 cm depth decreased by 0.053 and 0.039 in August, 2019 and 0.058 and 0.032 in August, 2020, respectively. For the same interval of time the SWCs at 70 and 100 cm depths of soil only decreased by 0.014 and 0.006 in 2019 and by 0.009 and 0.022 in 2020, respectively. These differences may be the fact that water at the top part continued to move down the profile. The middle soil was leaching water to the soil below it but at the same time it was receiving water from the soil above it, which resulted in the slower rates of SWCs decrease at 70 and 100cm depth than those at the top part (20 and 30cm). The similar patterns of changes in SWCs were observed in irrigated agricultural field of a desert oasis in Northwest China (Li et al., 2018).
Owing to the root water uptake in 2020, the SWPs variation mainly occurred in the upper 50 cm layer during the jointing period because of evapotranspiration. The SWPs at 20, 30 and 50cm showed a remarkable decrease with time, ranging from the average value of -204.3cm H2O on 28 June to -336.5 cm H2O on 6 July. During the tasseling period, the SWPs at 70 and 100 cm were found to exhibit abruptly lower values than those on the jointing period. The SWPs of 70 and 100cm soil depths continuously decreased with average value from -213.4cm H2O on 6 July to -400.3 cm H2O on 1August. The results indicate that some of the maize root system has reached the depth of 100cm during the tasseling periods due to the increased water consumption. The maximum root depth was similar to that in Beijing described in the related research (Ma and Song, 2016).
Since the soil water at 0-50 cm for bare ground surface is mainly affected by precipitation, evaporation, the SWP and SWC at these depths have relatively major changes with time (Figs. 3 and 4). The isotopic compositions of the soil water samples collected from 20,30 and 50 cm depths vary more diverse with the isotopic values of rain than those from other depths. Those implied that ordinary rainfall events and even continuous rain with 59.2 mm rain rarely affected the soil water in the deeper layers.

Estimation of soil water storage change and recharge to groundwater

For the observation period, the soil water storage, evapotranspiration /soil evaporation, infiltration (irrigation plus precipitation), deep percolation and groundwater capillary rise in the different land cover types between May and September was calculated using the water balance Eq. (1), and the results are presented in Table 5.The soil water storage change differed for bare ground in 2019 and for maize in 2020 about the amount and temporal development. The soil water storage (SWS)in 2019 firstly decreased from 7 may to 11 June (S1 and S2 stage) and then increased from 12 June to 5 September (S3, S4 and S5 stage) and lastly decreased from 6 to 18 September (S6 stage). The increasing SWS may result from the downward movement of the soil water from infiltration of irrigation/ precipitation and the depletions of SWS could be due to the soil evaporation. The infiltration was the main water input component. The highest (183.3 mm) and lowest (7.8 mm) volumes of water occurred at S5 and S1 stage, respectively. The capillary rise from groundwater accounted for only 2.2% of the total inputs in the balance, which varied temporally according to groundwater level. In terms of outputs, the deep percolation (312.0 mm) was the main component throughout the entire study period, accounting for 57.4% of the total outputs. The lowest value was recorded at S1 stage (3.1 mm, 0.6% of outputs), while the highest occurred at S5 stage (183.3 mm, 24.1% of outputs). As expected, deep percolation increased with the expansion of infiltration, from 66.1mm at S3 stage to 131.0 mm at S5 stage. The soil evaporation (231.2mm) corresponded to 42.6% of the outputs in the entire study period of 2019. Although the highest soil evaporation (69.3 mm) was recorded at S4 stage, the infiltration was not largest at this stage. The data shows that the soil evaporation is not only affected by the SWS, but also by the other factors, such as temperature, wind speed etc.
Temporal dynamics of the SWS was different for maize in 2020 compared with bare ground in 2019. Field water balance components over the growing season status was calculated (Table 5). The SWS decreases during the three periods (at the sowing, tasseling, and filling stage) due to the excess of evapotranspiration relative to the irrigation/precipitation infiltration. The largest depletions of the SWS (-58.1mm) could be observed at tasseling stage, which is likely caused by increasing maize water use during the growing season. The infiltration and capillary rise were 535.9mm and 8.7mm, respectively, which were similar with that in 2019. However, the highest value of infiltration was 150.1 mm recorded at seedling stage which was smaller than that at S5 stage in 2019. The total deep percolation was 156.6mm, accounting for 29.2% of the total infiltration precipitation or irrigation. The deep percolation was lowest (3.3mm) at the sowing stage and highest (41.4mm) at tasseling stage. The ET in the maize field was larger than deep percolation, which suggests that the water lost by ET in the irrigated areas was the greatest output component in the water balance. The total ET for maize was 431.3mm, which was very similar to that found by other researchers (Ren et al., 2016; Xu et al., 2013). In addition, the roots of maize at tasseling stage were mainly distributed in the depth range of 0–100 cm, where soil water consumption intensity was the highest over observation period. Therefore, the maize development and grain yield will consume more soil water.
As already mentioned, the depth of infiltration from precipitation was approximately 50 cm, indicating that the infiltration of precipitation cannot be attributed to groundwater recharge. However, the groundwater level response following surface irrigation was obtained. Combined with the analysis of the water table fluctuations, it can be reasonably concluded that the groundwater may be recharged by the lateral groundwater flow leaked from the ditch and the infiltration of surface irrigation. The fast leakage of ditch resulted in maximal groundwater level response following most irrigations. For example, the water table response occurred on 1 July 2019, with a maximum rise of 16 cm to the beginning of irrigation (Fig. 2). The water level before the start of irrigation was about 360 cm below ground level and it was observed that a significant water level rise (114cm) started about 22 h after the end of irrigation. The observed rise rates of the water level are approximately 9.5cm/h. However, the small deep percolation was only 30 mm during the irrigation which is the smallest recharge into groundwater. It can be speculated that the recharge from the leakage of ditch is indeed responsible for the groundwater level rise on this basis.