Urbanization significantly impacts hydrological processes and the water environment in the urban area. Studies on soil moisture are necessary to explore the hydrological characteristics against the background of urbanization. A regional grassland with Bermudagrass in Yangzhou City, China, was adopted as the study area. An analysis of its response characteristics to rainfall infiltration was carried out for different rainfall events. The Penman–Monteith equation was used to estimate the hourly evapotranspiration (ET) of the soil layers at depths of 0–30 cm and 0–10 cm. The results indicate that the response time of the soil water content in the root layer (0–10 cm) decreases with the decrease of the soil water content in the topsoil. The rate of the increase of the soil water content increases as the rainfall intensity increases in the state of unsaturation. The soil water in the root layer provided more than 70% of the total ET. The Nash efficiency coefficient of the simulation results of the soil water content at different depths obtained using the Hydrus-1D model was more than 0.75. The rationality of the results for the different rainfall events and the infiltration depth were verified via numerical simulations using the Hydrus-1D model.

  • Studies on soil moisture are necessary to explore hydrological characteristics in an urban region.

  • A regional Bermudagrass in Yangzhou City was adopted as the study area.

  • Main response characteristics of soil moisture to rainfall were clarified.

  • ET mainly provided by the soil water in the root layer was revealed.

  • The rationality for rainfall events and the infiltration depth were verified using the Hydrus-1D.

Graphical Abstract

Graphical Abstract
Graphical Abstract

Urbanization leads to widely occurring land use and land cover changes (Zhang et al. 2020; Imran et al. 2021; Mohammad et al. 2021). Accelerated urbanization continues to convert natural land types to impervious surfaces, resulting in serious environmental impacts and affecting the growth of urban plants (Song et al. 2015). The urbanization process alters the hydrological performance of an area (Armson et al. 2013). Urbanization causes a range of environmental challenges as a direct result of the biochemical and physical changes to the hydrological systems (Fletcher et al. 2013; Zang et al. 2019). Compared with the original landform units, decreases in the water retention function of various artificially disturbed landform units due to urbanization activities are the main factor causing urban water and soil losses and the aggravation of urban floods under certain rainfall conditions and in specially designed drainage networks (Baek et al. 2015; Shi et al. 2016). Urban grasslands are expanding rapidly with urbanization, which is expected to increase at unprecedented rates in the near future. Compared with impervious areas, grassland is able to effectively delay runoff (Liu et al. 2020a). The large and increasing area of urban grasslands and their impact on water justify the need for a better understanding of these areas (Duran et al. 2013). Grassland is an indispensable part of the urban green space ecosystem. It plays an important role in improving the regional ecological environment, regulating the hydrological cycle, and reducing the loss of soil and water (Livesley et al. 2010; Xiong et al. 2014; Xu & Cheng 2019; Yang et al. 2020). Soil moisture plays a critical role in land surface–plant–atmosphere interactions, and it directly impacts food security, human health, and ecosystem functions (Huang & Shao 2019). Evapotranspiration (ET) is an important hydrological process in the water cycle and plays a key role in the energy budget and water balance of the earth–atmosphere system (Zhang & Shen 2007; Zhao et al. 2018). It is important to study the changes in the soil moisture and ET of grassland vegetation under the background of urbanization to improve the urban ecological environment and enhance the construction level of a sponge city. Many studies indicated that rainfall affects the change of soil moisture. Wiekenkamp et al. (2016) conducted research on the spatial and temporal occurrences of the preferential flow in a forested headwater catchment using the preferential flow model and found that rainfall has a significant influence on soil moisture. Liu et al. (2020b) analyzed the variation characteristics of soil moisture parameters during rainfall and proposed that the amount of rainfall is the most prominent rainfall feature controlling the soil moisture response. Chen et al. (2020) conducted research on the response of the soil moisture to rainfall events in black locust plantations at different stages of restoration in the hilly-gully area of the Loess Plateau, China, and found that rainfall infiltration mainly occurred at soil depths of 0–60 cm. The rainfall infiltration was mainly jointly influenced by the rainfall attributes and soil properties. In contrast, there are relatively few studies on the response characteristics of grassland to rainfall in urban areas in China. Determining the response characteristics of soil moisture to rainfall in urban grassland can provide a basis for the studies on soil moisture movement in grassland vegetation against the background of urbanization. This is of great significance for developing the grassland vegetation, conserving water resources, and improving the hydro-ecological environment in urban areas. In this study, a regional urban grassland was adopted as the study location to reveal the response characteristics of the soil moisture in grassland vegetation areas to rainfall against the background of urbanization. Analysis of the infiltration characteristics during different rainfall events was carried out, the ET over the grassland in the study area was calculated and evaluated, and the infiltration depths of the different rainfall events were verified via numerical simulation using the Hydrus-1D model. The results of this study provide a scientific basis for further studies of soil moisture in grasslands during urbanization and the improvement of the urban ecological environment based on the development of grasslands.

General setting of the study area

Yangzhou City is located in the middle of Jiangsu Province on the southern part of the Jianghuai Plain. The study area has a subtropical humid monsoon climate. The annual average temperature is 16.1 °C. The annual average rainfall is about 1,000 mm, and 67% of the total rainfall is concentrated in the main rainy season from May to September (Zhou et al. 2019). Yangzhou City is one of the national ecological garden sites in China. There are many green spaces and parks in this urban area, with a green coverage rate of 44.03%, which plays an important role in improving the water circulation and reducing the waterlogging in the Yangzhou urban area.

A regional artificial lawn on the Yangzijin Campus of Yangzhou University was chosen as the study area, which is located in the southwestern part of Yangzhou City (the geographical coordinates of the control point are 32°20′58″ N, 119°23′51″ E; area of 340 m2). The study area is covered by a single species of grass (Bermudagrass), and the coverage is nearly 100% (Figure 1). Bermudagrass is a perennial warm season herb, with a main growth period from May to September. It is drought resistant, weed resistant, and very adaptable, with root depths of 8–10 cm. As a primarily green grass, Bermudagrass has been widely planted in parks, communities, and schools in southern China.
Figure 1

Grass cover (partial view) and field observations at the study area.

Figure 1

Grass cover (partial view) and field observations at the study area.

Close modal

According to the field survey, the 0–60 cm soil depth in the study area consists of mainly silty loam. The groundwater (phreatic water) level varies with the seasons and is about 2–5 m from the ground surface. There are many buildings and impervious pavement areas around the study area.

Data acquisition

An automatic weather station (mode: U30-nrc-10-s100-000; Onset Company, USA) was set up in the study area to record the hourly temperature, rainfall, solar radiation, wind speed, and relative humidity. The elevation of the observation point was 14 m, and the equipment height was 2 m (Figure 1(b)). The soil water contents were measured in two sites using soil moisture meters (mode: H21-002; Onset Company, USA), the observation depths of one point (P1) were 10, 25, and 40 cm, while those of the other point (P2) were 5, 15, 30, and 60 cm.

To evaluate the response of the soil moisture in the root layer of the Bermudagrass lawn to different rainfall processes, a simple soil experiment was carried out according to the weather forecasting information. The soil water content of the topsoil (0–5.5 cm) at point P1 was sampled and weighed before each rainfall event, and then it was weighed after drying to determine the initial water content of the topsoil. In addition, a tipping bucket rain gauge (mode: 7852M-L10, Onset Company, USA) was used to measure the rainfall at 10-min intervals, while the soil water content at point P1 (10, 30, and 60 cm) was measured every 10 min during the rainfall observations.

Additionally, the soil samples were collected from different depths in the study area. A laser particle size analyzer (model: Mastersizer 3000E; Malvern Company, UK) was used to measure the particle compositions of the soil samples, and the particle size distributions (clay, sand, and silt contents) were obtained.

ET calculation

The Penman–Monteith (P–M) equation (Equation (1)) was used to calculate the ET over the grassland in the study area since it is widely used for the calculation of the ET over various vegetation types (Longobardi & Villani 2013; Hadi & Farah 2018; Djaman et al. 2019).
(1)
where λ is the latent heat of vaporization (MJ·kg−1); cp is the specific heat of the air at a constant pressure (1.0 × 10−3MJ · kg−1· K−1); Δ is the slope of the saturation vapor-pressure at the air temperature (kPa·K−1); Rn is the net radiation (MJ · m−2 · h−1); G is the soil heat flux (MJ · m−2 · h−1); ρ is the air density at constant pressure (kg · m−3); γ is the psychometric constant (kPa · K−1); es is the saturated vapor-pressure at the air temperature (kPa); ea is the actual vapor-pressure (kPa); and ra and rs are the aerodynamic resistance and the surface resistance, respectively (s·m−1).
The methods of calculating each factor in Equation (1) have been described by Zhou et al. (2019). The surface resistance (rs), which is controlled by the soil water content of the 0–30 cm soil layer (rs30) and of the root layer (depth of about 10 cm) was calibrated following the method of Kimura et al. (2005) and Zhou et al. (2019), and the results were determined using Equations (2) and (3), respectively.
(2)
(3)
where θ30 is the average soil water content in the 0–30 cm soil layer (cm3·cm−3); and θ10 is the soil water content at a depth of 10 cm (cm3·cm−3).

Hydrus-1D model description

Basic equations

Simulations of the soil water content at different depths were performed using the Hydrus-1D model to validate the infiltration depths of the different rainfall events. The Richards equation is used by the Hydrus-1D model to numerically solve the soil moisture movement in variably saturated porous media (Fairouz et al. 2020; Narjary et al. 2020). The Richards equation (Equation (4)), with the water content as the dependent variable, was used to construct the water transport model.
(4)
where θ is the soil water content (cm3·cm−3); t is the time factor (h); z is the vertical distance from the ground (soil depth), with the coordinate system being positive downward (cm); D(θ) is the soil water diffusivity (cm2·h−1); K(θ) is the unsaturated hydraulic conductivity (cm·h−1); S is the source (or sink) of the soil water, which represents the water absorption rate of the crop roots (cm·h−1); Ks is the saturated hydraulic conductivity (cm·h−1); θe, θs, and θr are the effective, saturated, and residual water contents, respectively (cm3·cm−3); α and n are the empirical shape parameters, with m = l − 1/n; l is a pore connectivity parameter; and h is the soil matrix potential (cm).

Definite solution conditions

The initial condition was set as the observed value of the soil water content at the beginning of the calculation. The upper boundary condition was set as the atmospheric boundary condition with the surface layer, which includes the hourly (1 h series) rainfall and potential evapotranspiration (ET0). The hourly ET0 was estimated using the PM model, which is recommended by the FAO 56 (Allen et al. 1998). The lower boundary was located in the unsaturated zone because it does not reach the phreatic layer, so it was set as a free drainage boundary. The definite solution conditions are expressed by Equation (5).
(5)
where θ0 is the initial soil water content (cm3·cm−3); q0 (t) is the soil water flux (cm·d−1); L is the vertical depth of the lower boundary (cm); and θL(t) is the soil water content at the lower boundary (cm3·cm−3).

Parameter calibration

According to the screening results of the soil particle size obtained from the particle composition of the soil samples (Table 1), the Rosetta module, which is based on a neural network, was used to preliminarily determine the saturated water content θs, the residual water content θr, and the saturated hydraulic conductivity Ks (Li et al. 2015). Based on this, comparisons of the Hydrus-1D calculated results and the measured soil water contents at various depths (5, 15, 30, and 60 cm) were performed. The adjustment of each parameter value and repeated model calculations were carried out to reduce the difference between the simulated results and the measured results to achieve the calibration of the parameters (Table 1).

Table 1

The measured data of soil physical properties and main parameters obtained by combined using Rosetta and model calculation

Soil depth (cm)Sand grain % 50–2,000 μmSilt % 2–50 μmClay % <2 μmθs cm3·cm−3θr cm3·cm−3α cm−1nKs cm·min–1
0–10 27.10 65.71 7.19 0.35 0.045 0.0047 1.23 2.08 0.5 
10–20 21.78 69.76 8.48 0.37 0.051 0.0046 1.23 2.33 0.5 
20–40 17.72 74.05 8.24 0.36 0.056 0.0048 1.33 1.85 0.5 
40–60 19.33 71.23 9.45 0.37 0.053 0.0045 1.42 1.43 0.5 
Soil depth (cm)Sand grain % 50–2,000 μmSilt % 2–50 μmClay % <2 μmθs cm3·cm−3θr cm3·cm−3α cm−1nKs cm·min–1
0–10 27.10 65.71 7.19 0.35 0.045 0.0047 1.23 2.08 0.5 
10–20 21.78 69.76 8.48 0.37 0.051 0.0046 1.23 2.33 0.5 
20–40 17.72 74.05 8.24 0.36 0.056 0.0048 1.33 1.85 0.5 
40–60 19.33 71.23 9.45 0.37 0.053 0.0045 1.42 1.43 0.5 

Response time of the soil moisture in the root zone to rainfall

Six periods with rainfall events from April to June in 2019 were chosen to analyze the changes in the soil water content during the different rainfall processes and to estimate the response time of the soil moisture in the root layer to the rainfall. The response time of the soil moisture in the root layer refers to the time from the beginning of the rainfall to the beginning of the soil water content change at a depth of 10 cm. The soil water content of the topsoil (5.5 cm) and the response time at a soil depth of 10 cm for each rainfall event are shown in Table 2.

Table 2

Response times of the soil water content to the different rainfall events in the grass root zone at a depth of 10 cm

No.Soil water content before rainfall (cm3·cm−3)
Response time (min)
Topsoil/5.5 cmRoot zone/10 cm
0.215 0.322 620 < t ≤ 630 
0.021 0.201 t ≤ 10 
0.023 0.195 t ≤ 10 
0.045 0.254 t ≤ 10 
0.187 0.329 50 < t ≤ 60 
0.131 0.310 30 < t ≤ 40 
No.Soil water content before rainfall (cm3·cm−3)
Response time (min)
Topsoil/5.5 cmRoot zone/10 cm
0.215 0.322 620 < t ≤ 630 
0.021 0.201 t ≤ 10 
0.023 0.195 t ≤ 10 
0.045 0.254 t ≤ 10 
0.187 0.329 50 < t ≤ 60 
0.131 0.310 30 < t ≤ 40 

Note: The response time is expressed by a time interval of 10 min as the unit time of rainfall and soil water content is set to 10 min.

The temporal process of each rainfall event and the changes in the soil water content at different depths are shown in Figure 2. From 13:00 on April 28th to 13:00 on April 30th, the amount of rainfall was 6.4 mm, with a maximum intensity of 0.42 mm · min−1. The response time was more than 10 h due to the frequent rainfall interruptions, the short durations, and the small amount of periodic rainfall (Figure 2(a)). Moreover, as the Bermudagrass lawn had a high coverage and dense vegetation interception, ET consumed most of the rainfall, and there was no effective infiltration within a few hours after the rainfall. After the soil water content at a depth of 10 cm (root layer) reached the peak, the soil water content at a depth of 25 cm increased slowly, while the soil water content at a depth of 40 cm remained relatively stable (Figure 2(a)).
Figure 2

Changes of the soil water content under the different rainfall events. (a) April 28th 13:00 to April 30th 13:00 (b) May 25th 21:00 to May 27th 17:00 (c) June 5th 21:00 to June 7th 1:00 (d) June 17th 18:00 to June 18th 22:00 (e) June 20th 11:00 to June 21st 9:00 (f) June 28th 22:00 to June 29th 7:00.

Figure 2

Changes of the soil water content under the different rainfall events. (a) April 28th 13:00 to April 30th 13:00 (b) May 25th 21:00 to May 27th 17:00 (c) June 5th 21:00 to June 7th 1:00 (d) June 17th 18:00 to June 18th 22:00 (e) June 20th 11:00 to June 21st 9:00 (f) June 28th 22:00 to June 29th 7:00.

Close modal

The change processes of the soil water content at depths of 10, 25, and 40 cm from 21:00 on May 25th to 17:00 on May 27th are shown in Figure 2(b). A relatively concentrated rainfall event occurred on May 26th, with a rainfall amount of 12.6 mm and a maximum intensity of 0.20 mm · min−1. The response time of root layer to the rainfall occurred within 10 min, and the range of the increase in the soil water content in the root layer was significantly affected by the rainfall intensity. The soil water contents at 25 and 40 cm remained stable. The main reasons for these results are that the original soil water content of the topsoil was low before the beginning of the rainfall event (Table 2), the soil water content of the root zone did not attain saturation after the end of the rainfall event, and the soil water content below 25 cm was not recharged by the infiltration.

Figure 2(c) shows that a rainfall event occurred from 22:00 on June 5th to 11:00 on June 6th, with a rainfall amount of 30.4 mm and a maximum intensity of 0.46 mm · min−1. The water content of the topsoil was low before the rainfall event, and the response time to the rainfall was less than 10 min. The soil water content of the root zone increased slowly when the rainfall intensity was small and the rate of increase improved with the increasing rainfall intensity. The soil water content of the root zone increased significantly during 5:50–6:20 and 9:10–11:20 on June 6th since the maximum rainfall intensity reached 0.20 and 0.46 mm 10 min−1 in these two periods, respectively, which indicates that the rainfall intensity greatly affects the change in the soil water content of the root zone. The soil water content at 25 cm began to increase slowly from the time of the maximum rainfall intensity, while the soil water content at 40 cm remained stable.

Figure 2(d) shows the temporal changes in the soil water content at depths of 10, 25, and 40 cm from 18:00 on June 17th to 22:00 on June 18th. A rainfall event occurred from 20:00 on June 17th to 9:00 on June 18th, with a rainfall amount of 14 mm and a maximum intensity of 0.14 mm · min−1. The response time of the water content at a depth of 10 cm to the rainfall occurred within 10 min, and the soil water content at 25 cm increased slightly when the rainfall intensity reached the maximum, while the soil water content at 40 cm remained stable.

A relatively concentrated rainfall event occurred on June 20th (14:00–20:00), with a rainfall amount of 8 mm and a maximum intensity of 0.08 mm · min−1. The response time of the water content at 10 cm to the rainfall was 50–60 min, and the soil water contents at depths of 25 and 40 cm remained stable during this rainfall event (Figure 2(e)).

Figure 2(f) shows the temporal changes in the soil water content at depths of 10, 25, and 40 cm from 18:00 on June 17th to 22:00 on June 18th, and a rainfall event occurred from 22:00 on June 28th to 7:00 on June 29th, with a rainfall amount of 14 mm and a maximum intensity of 0.48 mm · min−1. The rainfall duration was relatively short, and the intensity was comparatively large. The response time of the water content at 10 cm to the rainfall was 30–40 min. The main reasons for these results are that the initial soil water contents of the topsoil and the root zone were relatively high before the rainfall event (Table 2), which resulted in a decreased infiltration rate.

To summarize the above analysis, the response time of the soil water content of the root zone to rainfall was affected by the actual rainfall process and the underlying surface conditions. In general, as the rainfall intensity increased, the response time of the soil water content of the root zone to the rainfall decreased. For similar rainfall intensities, the lower initial water content of the topsoil causes a shorter response time in the root layer to rainfall and a faster infiltration rate in the topsoil. The main reason for this is that the lower water content of the soil results in a lower soil water potential, so the suction of the water increases and the infiltration rate improves. From the beginning of the rainfall to a short time after the end of the rainfall, the water content below 25 cm could not receive or only received a small amount of effective infiltration recharge, while the soil water content at 40 cm remained relatively stable, which indicates that the grassland vegetation in the study area has a strong interception function. Before the soil water content of the root zone reaches saturation, the soil water content is significantly impacted by the rainfall intensity; the greater the rainfall intensity, the faster the soil water content increases. Yang et al. (2008) obtained similar results in another study conducted on the Loess Plateau, China. However, due to the limitations of the observed data, the quantitative evaluation of the relationship between the change in the rainfall intensity and the increment of the soil water content of the root zone was not possible in the current research.

Evapotranspiration

The hourly ET of the 0–30 cm soil layer (ET30) and of the root zone (ET10) were estimated using the calibrated surface resistances (rs30 and rs10, respectively). The estimated period was from July 2018 to June 2019, which included the main growth period of the grass, and the results represent the change in the ET over the grassland (Bermudagrass lawn). ET30 and ET10 are shown in Figure 3. The maximum ET10 and ET30 are 1.04 and 1.11 mm, respectively. ET30 is higher than ET10 at any given time, whereas the temporal changes in ET10 and ET30 are similar, and the correlation coefficient is 0.99. The ET changes significantly with the season. The ET exhibits comparatively higher values in the main growth period of the grass (from June to September), while it exhibits relatively low levels in the non-main growth period (January, February, and December). The cumulative ET30 and ET10 values are 396 and 309 mm, respectively, which account for 39.6 and 30.9% of the rainfall in the same period, respectively. The cumulative ET10 is 78% that of ET30, which indicates that the soil water in the root zone provides the main part of the water consumed via ET.
Figure 3

Calculated results of hourly ET30 and ET10.

Figure 3

Calculated results of hourly ET30 and ET10.

Close modal
Figure 4 shows the randomly selected results of the hourly ET10 on 4 non-rainfall days in different seasons. The maximum difference in the ET10 between two adjacent hours was 0.10 mm on October 1, 2018, and the daily ET10 was 0.41 mm. On December 30, 2018, ET10 did not fluctuate significantly; its maximum difference between two adjacent hours was 0.03 mm, and the daily accumulated value was 0.28 cm. On April 7, 2019, ET10 fluctuated within different ranges, the maximum difference between two adjacent hours was 0.09 mm, and the daily accumulated value was 0.49 mm. ET10 fluctuated frequently on June 30, 2019, the maximum difference between two adjacent hours was 0.37 mm, and the daily accumulated value was 2.91 mm. The daily ET10 generally increased after a certain time in the morning, gradually decreased after reaching the peak, and approached 0 after a certain time in the evening. When the meteorological factors (such as the temperature and solar radiation) that affect ET10 changed significantly within a certain period in the daytime, ET10 fluctuated upward or downward. During the main growth period from June to September, the soil water content and the main meteorological factors, which significantly affect the ET, mostly remained at their comparatively high annual levels, leading to a relatively large accumulated daily ET10. The daily change in ET10 in the winter was smaller than that in the other seasons.
Figure 4

Daily change of ET10.

Figure 4

Daily change of ET10.

Close modal

Amount of rainfall and infiltration depth

The eight rainfall events and the observed soil water contents on P2 (at depths of 5, 15, 30, and 60 cm) during the same periods were chosen to explore the relationship between the amount of rainfall and the infiltration depth. The grades of the eight rainfall events were classified (Table 3) according to the classification standard of the China Meteorological Administration.

Table 3

Index of the eight rainfall events and infiltration depth

Grade of rainfall eventThe periodic timeRainfall concentrated timeRainfall amount (mm)Infiltration depth (cm)
Light rain 10.08.2018–11.10.2018 10.09 14:00–16:00 4.8 <5 
30.05.2018–02.06.2018 05.30 22:00–05.31 7:00 8.9 5–15 
Moderate rain 21.07.2018–24.07.2018 07.22 11:00–07.23 3:00 17.8 15–30 
04.04.2018–07.04.2018 04.05 12:00–23:00 24.0 
Heavy rain 12.08.2018–15.08.2018 08.13 00:00–17:00 37.6 30–60 
05.06.2019–08.06.2019 06.06 1:00–14:00 41.2 >60 
Rainstorm 09.08.2019–12.08.2019 08.10 2:00–08.11 1:00 55.0 >60 
23.05.2018–26.05.2018 05.24 22:00–05.25 20:00 107.0 >60 
Grade of rainfall eventThe periodic timeRainfall concentrated timeRainfall amount (mm)Infiltration depth (cm)
Light rain 10.08.2018–11.10.2018 10.09 14:00–16:00 4.8 <5 
30.05.2018–02.06.2018 05.30 22:00–05.31 7:00 8.9 5–15 
Moderate rain 21.07.2018–24.07.2018 07.22 11:00–07.23 3:00 17.8 15–30 
04.04.2018–07.04.2018 04.05 12:00–23:00 24.0 
Heavy rain 12.08.2018–15.08.2018 08.13 00:00–17:00 37.6 30–60 
05.06.2019–08.06.2019 06.06 1:00–14:00 41.2 >60 
Rainstorm 09.08.2019–12.08.2019 08.10 2:00–08.11 1:00 55.0 >60 
23.05.2018–26.05.2018 05.24 22:00–05.25 20:00 107.0 >60 

The maximum duration of the selected rainfall event was 23 h. To ensure sufficient infiltration during each rainfall event, a periodic time of 4 days was adopted when analyzing each rainfall event and the subsequent infiltration. The amount of rainfall and the infiltration depth during each period are shown in Table 3. The changes in the soil water content at different depths are shown in Figure 5.
Figure 5

The change of soil water content at the depths of 5, 15, 30, and 60 cm during the selected eight periods: (a) 5 cm, (b) 15 cm, (c) 30 cm, and (d) 60 cm.

Figure 5

The change of soil water content at the depths of 5, 15, 30, and 60 cm during the selected eight periods: (a) 5 cm, (b) 15 cm, (c) 30 cm, and (d) 60 cm.

Close modal

Figure 5(a) shows the changes in the soil water contents at the four depths (5, 15 30, and 60 cm) during the period from October 8th to October 11th in 2018. The rainfall was mainly concentrated on October 9th, and the amount of rainfall was 4.8 mm (Table 3). The soil water contents at the four depths remained stable, that is, the rainwater infiltration did not reach a depth of 5 cm. The main reason for this is that the Bermudagrass lawn vegetation has a strong interception function so effective infiltration did not occur during this rainfall event. A relatively concentrated rainfall event occurred from 22:00 on May 30th to 7:00 on May 31st, with a rainfall amount of 8.9 mm. The soil water content at a depth of 5 cm increased significantly due to rainwater infiltration. It reached its peak on May 31, gradually decreased until June 2, and then returned to the initial state observed on May 30th. The soil water contents at 15, 30, and 60 cm remained stable during the 4 days (2018/05/30–06/02) (Figure 5(b)). The above two rainfall events were light rainfall events. The amount of rainfall during the second rainfall event (8.9 mm) was greater than that during the first event (4.8 mm), and its infiltration depth was also greater than that of the first event.

Two moderate rainfall events occurred from July 22nd to July 23rd and on April 5, 2018, and the amounts of the rainfall during these two events were 17.8 and 24.0 mm, respectively. The soil water contents at 5 and 15 cm increased significantly due to the infiltration caused by the two rainfall events. The changes in the water content at depths of 5 and 15 cm exhibited similar temporal patterns, that is, they gradually decreased after reaching their peak. The range of the change in the soil water content at 5 cm was greater than that at 15 cm, whereas the soil water contents at 30 and 60 cm remained stable. And the infiltration depth caused by moderate rain generally was less than 30 cm (Figure 5(c) and 5(d)).

Two heavy rainfall events occurred on August 13, 2018, and June 6, 2019, with rainfall amounts of 37.6 and 41.2 mm, respectively. The soil water contents at 5, 15, and 30 cm increased to different degrees due to infiltration. The soil water content at 60 cm remained relatively stable during the rainfall event and after the end of the rainfall on August 13, 2018, which indicates that the rainfall infiltration did not reach 60 cm (Figure 5(e)). Due to the infiltration effect of the rainfall event on June 6, 2019, the soil water content at 60 cm increased slightly (Figure 5(f)), that is, the infiltration due to this rainfall event reached 60 cm. The two heavy rainfall events both led to increases in the soil water contents at depths of 5, 15, and 30 cm, but the ranges of these increases were different. Generally, the range of the change in the soil water content decreased with increasing depth. (Figure 5(e) and 5(f)). The results also show that the infiltration depth of the heavy rainfall events exceeded 30 cm, and some reached 60 cm (Figure 5(e) and 5(f)).

Two rainstorms occurred from May 24th to 25, 2018, and from August 10 to 11, 2019, with rainfall amounts of 107 and 55 mm, respectively. The soil water contents at 5, 15, 30, and 60 cm visibly increased, and the infiltration depth exceeded 60 cm. However, the ranges of the changes in the soil water content and the water content processes were different at different depths. (Figure 5(g) and 5(h)).

The analysis results of the randomly selected rainfall events (including but not limited to the above described eight rainfalls) and the changes in the soil water contents during the same periods indicate that the infiltration depth generally varies with the amount of rainfall. Of the 103 rainfall events that occurred during the study period, 77 light rainfall events had infiltration depths of less than 15 cm, and more than 80% had infiltration depths of even less than 5 cm. Most of the 20 moderate rainfall events had infiltration depths mostly less than 15 cm, and only three of them had infiltration depths of 30 cm (<60 cm). For the four heavy rainfall events, the infiltration depths exceeded 30 cm, and only one of them reached 60 cm. Besides, the infiltration depths for two rainstorms exceeded 60 cm. The results of the linear regression analysis showed a high goodness of fitting (R2 = 0.705) between the rainfall amount and corresponding infiltration depths for the 103 rainfall events (Figure 6). A larger amount of rainfall generally results in a greater infiltration depth. In the study area (Bermudagrass lawn), the infiltration depths of the light rainfall events were generally less than 15 cm, and those of the moderate rainfall events were mostly less than 30 cm; while the infiltration depths of the heavy rainfall events were generally greater than 30 cm (with some greater than 60 cm), and the infiltration depth of the rainstorm was even greater (>60 cm). Such results are basically consistent with those of Wu et al. (2018), which were achieved using a similar research method.
Figure 6

Linear regression analysis between rainfall amount and infiltration depths. In Figure 6, h refers to the infiltration depth (cm). Though water infiltration in the soil was continuous, this study only measured soil water content at 5, 15, 30, and 60 cm on one of the two points (P2). In other words, if the ultimate infiltration depth for a certain rainfall event was between two measured depths (such as deeper than 5 cm but was above 15 cm), then it would not be measured. Also, the infiltration between two adjacent measured depths was not clear either.

Figure 6

Linear regression analysis between rainfall amount and infiltration depths. In Figure 6, h refers to the infiltration depth (cm). Though water infiltration in the soil was continuous, this study only measured soil water content at 5, 15, 30, and 60 cm on one of the two points (P2). In other words, if the ultimate infiltration depth for a certain rainfall event was between two measured depths (such as deeper than 5 cm but was above 15 cm), then it would not be measured. Also, the infiltration between two adjacent measured depths was not clear either.

Close modal

There are a number of factors that affect infiltration such as topography, soil properties, and vegetation (Leonard & Andrieux 1998; Ribolzi et al. 2011; Huang et al. 2013; Nie et al. 2017). The topography of the study area is flat, with large areas having almost no slope (Figure 1(b)). The vegetation within the study area, as a regional artificial lawn, is a single species of grass (Bermudagrass) with uniform distribution. According to the screening results of soil particle composition experiment, the soil has good homogeneity, and the soil physical properties are in the similar conditions (Table 1). Therefore, the response characteristics of soil moisture to rainfall obtained in this study, such as the response time of the soil moisture in the root zone to rainfall, and the relationship between the amount of rainfall and infiltration depth, can appropriately represent the infiltration characteristics of grassland in the study area.

Validation of the Hydrus-1D model

The root mean square error (RMSE, Equation (6)), the Nash efficiency coefficient (NSE, Equation (7)), Kling-Gupta Efficiency (KGE, Equation (8)) (Gupta et al. 2009), and Liu Efficiency (LME, Equation (9)) (Liu 2020) were combined and used to evaluate the simulation error of the Hydrus-1D model. The RMSE represents the average error between the simulated results and the measured results. When the RMSE is much closer to 0, the error is smaller, and the simulated results are more accurate. The NSE, KGE and LME characterize the model's efficiency. The model is more reliable when their values are much closer to 1.
(6)
(7)
(8)
(9)
(10)
(11)
(12)

In Equations (6)–(12), Si and Oi are the simulated value and the measured value, respectively; and devote the average of the simulated results and the measured results, respectively; r is the linear correlation coefficient between Si and Oi; α is a measure of relative variability in the simulated and measured values; β is the ratio between the average simulated value and the average measured value.

An hourly series simulation of the soil water content at depths of 5, 15, 30, and 60 cm from April 4, 2018, to September 19, 2019 (total 12,816 h), was carried out using the Hydrus-1D model. The results are shown in Figure 7. Figure 7 shows that the simulated results of the soil water content at different depths appropriately reproduce the changes in the measured results. However, slight differences exist in the simulated results of the soil water contents at different depths. The difference between the simulated result and the measured value for a depth of 5 cm is periodically larger than those of the other three depths (Figure 7). The RMSE of the simulated result for 5 cm is the highest (0.034 cm3·cm−3), the NSE (0.77), KGE (0.714) and LME (0.690) are the lowest among the errors for the four depths (Table 4), which indicates that the simulation error of the 5 cm value is the maximum. As the depth increases, the simulation's accuracy increases. Moreover, the KGE and LME for the depths of 15, 30, and 60 cm exceed 0.90 (Table 4). The soil water content at 5 cm is significantly affected by rainfall infiltration and ET, and the function that the Hydrus-1D model uses for the soil water simulation is limited to vertical movement and lacks a function for calculating the horizontal diffusion. In the process of calculating the vertical movement of the soil water, vegetation interception is regarded as part of the ET0, which cannot be accurately estimated during the rainfall process (Sutanto et al. 2012). Because the root depth of Bermudagrass is about 10 cm, as the depth increases, the effect of the ET and rainwater infiltration on the soil water content gradually decreases, and the accuracy of the simulation results gradually increases. Although the simulation error at 5 cm is the maximum among the four depths, the simulated results still have a comparatively high accuracy, with an NSE of approximately 0.80 (KGE exceeds 0.70, LME is about 0.70). The results of the error evaluation demonstrate that the Hydrus-1D model is applicable to the simulation of the soil water content in the study area, and the simulation results have a relatively high accuracy.
Table 4

Index of the simulation errors

Depth (cm)5153060
RMSE 0.034 0.022 0.015 0.012 
NSE 0.77 0.78 0.83 0.84 
KGE 0.714 0.903 0.930 0.945 
LME 0.690 0.906 0.932 0.967 
Depth (cm)5153060
RMSE 0.034 0.022 0.015 0.012 
NSE 0.77 0.78 0.83 0.84 
KGE 0.714 0.903 0.930 0.945 
LME 0.690 0.906 0.932 0.967 
Figure 7

The simulated results of soil water content for four measured depths of the study area (Sim: simulated result; Obs: observed value).

Figure 7

The simulated results of soil water content for four measured depths of the study area (Sim: simulated result; Obs: observed value).

Close modal

Model validation of infiltration depth

To verify the rationality of the results of the relationship between the rainfall amount and the infiltration depth for different rainfall events based on the analysis of the observation data (Figure 5), the observed results of eight rainfall events (Table 3) and the measured and simulated results of the soil water contents at different depths at corresponding times during each concentrated rainfall period were extracted and processed (Figure 8).
Figure 8

The change processes of soil water content at the four measured depths for eight rainfall events (Sim: simulated result; Obs: observed value). (a) 12:00–24:00 on October 9th 2018, (b) May 30th 17:00–May 31st 16:00, 2018, (c) July 22th 17:00–July 23rd 8:00, 2018, (d) April 5th 8:00–April 6th 7:00, 2018, (e) August 12th 21:00–August 13th 20:00, 2018, (f) 00:00–23:00 on June 6th, 2019, (g) August 10th 3:00–August 11th 8:00, 2019, and (h) May 24th 20:00–May 25th 1:00, 2018.

Figure 8

The change processes of soil water content at the four measured depths for eight rainfall events (Sim: simulated result; Obs: observed value). (a) 12:00–24:00 on October 9th 2018, (b) May 30th 17:00–May 31st 16:00, 2018, (c) July 22th 17:00–July 23rd 8:00, 2018, (d) April 5th 8:00–April 6th 7:00, 2018, (e) August 12th 21:00–August 13th 20:00, 2018, (f) 00:00–23:00 on June 6th, 2019, (g) August 10th 3:00–August 11th 8:00, 2019, and (h) May 24th 20:00–May 25th 1:00, 2018.

Close modal

Figure 8(a) shows that during the light rainfall event on October 9, 2018, and before and after it, the measured soil water content at 5 cm remained stable, while the simulated results increased slightly during the concentrated rainfall period. In the process of the Hydrus-1D calculations, the rainfall caused water input at the model's upper boundary and the soil water content of the model increased, which resulted in a slight difference between the measured and simulated soil water contents at 5 cm. This confirms that the Bermudagrass in the study area has a strong interception effect. Figure 8(b) shows the change in the soil water content during another light rainfall event from May 30th to 31st in 2018 and before and after it. The measured and simulated soil water contents at 5 cm increased by a similar increment, whereas the measured and simulated soil water contents at 15 cm remained stable. Based on this, the rationality of the relationship between the amount of rainfall and the infiltration depth for light rainfall (Figure 5(b)) is valid.

The simulated soil water contents for the two moderate rainfall events (Table 3) show that the soil water content at 30 cm did not change, but it increased at 15 cm (Figure 8(c) and 8(d)). The simulated results are consistent with the analyzed results of the relationship between moderate rainfall and the infiltration depth (Table 3).

The simulated soil water contents for heavy rainfall show that the infiltration depth was less 60 cm for the heavy rainfall on August 13, 2018 (Figure 8(e)), while the infiltration depth reached 60 cm due to the heavy rainfall on June 6, 2019 (Figure 8(f)). These results validate the relationship between heavy rainfall and the corresponding infiltration depth (Table 3).

The simulated results of the soil water contents for the two rainstorms (Table 3) demonstrate that the soil water contents at different depths increased visibly (Figure 8(g) and 8(h)), and the downward release of water occurred at the lower boundary (i.e., the soil water flux was not 0 at a depth of 60 cm) in the model's calculation process, which indicates that the infiltration depth exceeded 60 cm. Thus, the rationality of the infiltration depth caused by the rainstorm is verified.

According to the above analysis, the Hydrus-1D model can be used to accurately estimate the infiltration depths for different rainfall events in the absence of measured soil water content data for the study area.

In this study, a regional urban grassland site with a single type of grass (Bermudagrass) was chosen as the study object. The response characteristics of the soil water content to the rainfall, including the relationship between the rainfall intensity and the response time of the soil moisture in the root zone, the relationship between the amount of rainfall and the infiltration depth, and the change in the daily ET, were evaluated. The rationality of the relationship between the amount of rainfall and the infiltration depth obtained from the data analysis was validated via simulations using the Hydrus-1D model. The main conclusions reached are as follows.

  • The initial soil water content of the topsoil and the rainfall intensity significantly affect the response time of the soil water in the root zone to the rainfall. A lower topsoil water content and a higher rainfall intensity generally lead to a shorter response time of the soil moisture in the root zone (depth of 10 cm) to rainfall.

  • Generally, for the regional urban Bermudagrass in this study area, the rainfall events at different grades can lead to different infiltration depths. The infiltration depths under the light or moderate rainfall events are mostly less than 15 or 30 cm, respectively. While those under heavy rainfall events can exceed 30 cm, and those under rainstorms even exceed 60 cm.

  • The soil water moisture in the root zone provides the main part (>70%) of the soil water consumed by ET.

  • The Hydrus-1D model is applicable to the simulation of the soil water content in the study area. The infiltration depths of different rainfall events can be accurately estimated using the Hydrus-1D model when there is no measured soil water content data for the study area.

This study revealed the characteristics of the response of the soil water content to rainfall and the ET characteristics of Bermudagrass in an urban area to a certain extent. Based on these results, the characteristics and laws of soil water movement in urban Bermudagrass in the same area will be further carried out. The results of this study also provide a methodological reference for research on soil moisture movement characteristics in different types of grasslands in urban areas.

The authors gratefully acknowledge the financial support of the National Natural Science Foundation of China (NSFC: 41271046), the 2018 Young Blue Project of Yangzhou University, China, and the 2019 Science and Technology Innovation Cultivation Fund of Yangzhou University, China (2019cxj069).

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.

Allen
R. G.
,
Pereira
L. S.
,
Raes
D.
&
Smith
M.
1998
Crop Evapotranspiration: Guidelines for Computing Crop Water Requirements
.
FAO
,
Rome
, pp.
1
300
.
Armson
D.
,
Stringer
P.
&
Ennos
A. R.
2013
The effect of street trees and amenity grass on urban surface water runoff in Manchester, UK
.
Urban Forestry & Urban Greening
12
(
3
),
282
286
.
doi:10.1016/j.ufug.2013.04.001
.
Baek
S. S.
,
Choi
D. H.
,
Jung
J. W.
,
Lee
H. J.
,
Lee
H.
,
Yoon
K. S.
&
Cho
K. H.
2015
Optimizing low impact development (LID) for stormwater runoff treatment in urban area, Korea: experimental and modeling approach
.
Water Research
86
(
1
),
122
131
.
doi:10.1016/j.watres.2015.08.038
.
Chen
W. L.
,
Li
Z. S.
,
Jiao
L.
,
Wang
C.
,
Gao
J. Y.
&
Fu
B. J.
2020
Response of soil moisture to rainfall event in black locust plantations at different stages of restoration in hilly-gully area of the Loess Plateau, China
.
Chinese Geographical Science
30
(
3
),
427
445
.
doi:10.1007/s11769-020-1121-4
.
Djaman
K.
,
Sall
M.
,
Sow
A.
,
Manneh
B.
&
Irmak
S.
2019
Impact of air temperature and relative humidity measured over rice and grass canopies on Penman-Monteith reference evapotranspiration estimates
.
Journal of Irrigation and Drainage Engineering
145
(
1
),
1
8
.
06018008
.
Duran
J.
,
Rodriguez
A.
,
Morse
J. L.
&
Groffman
P. M.
2013
Winter climate change effects on soil C and N circles in urban grassland
.
Global Change Biology
19
,
2826
2837
.
doi:10.1111/gcb.12238
.
Fairouz
S.
,
Emna
G. E.
&
Rachida
B.
2020
Impact of rainfall structure and climate change on soil and groundwater salinization
.
Climatic Change
163
,
395
413
.
doi:10.1007/s10584-020-02789-0
.
Gupta
H. V.
,
Kling
H.
,
Yilmaz
K. K.
&
Martinez
G. F.
2009
Decomposition of the mean squared error and NSE performance criteria: implications for improving hydrological modelling
.
Journal of Hydrology
377
(
1–2
),
80
91
.
doi:10.1016/j.jhydrol.2009.08.003
.
Hadi
H. J.
&
Farah
A.
2018
Evaluating atmometer performance for estimating reference evapotranspiration in ventilated and unventilated greenhouses
.
Journal of Irrigation and Drainage Engineering
144
(
7
),
1
10
.
04018014
.
Huang
L.
&
Shao
M. A.
2019
Advances and perspectives on soil water research in China's Loess Plateau
.
Earth-Science Reviews
199
,
1
22
.
No.102962. doi:10.1016/j.earscirev.2019.102962
.
Imran
H. M.
,
Hossain
A.
,
Islam
A. K. M. S.
,
Rahman
A.
,
Bhuiyan
M. A.
,
Paul
S.
&
Alam
A.
2021
Impact of land cover changes on land surface temperature and human thermal comfort in Dhaka City of Bangladesh
.
Earth Systems and Environment
5
(
3
),
667
693
.
Kimura
R.
,
Fan
J.
,
Zhang
X. C.
,
Takayama
N.
,
Kamichika
M.
&
Matsuoka
N.
2005
Evapotranspiration over the grassland field in the Liudaogou Basin of the Loess Plateau, China
.
Acta Oecologica
29
(
1
),
45
53
.
Leonard
J.
&
Andrieux
P.
1998
Infiltration characteristics of soils in Mediterranean vineyards in Southern France
.
Catena
32
,
209
223
.
doi:10.1016/S0341-8162(98)00049-6
.
Li
H.
,
Yi
J.
,
Zhang
J.
,
Zhao
Y.
,
Si
B.
,
Hill
L. R.
,
Cui
L.
&
Liu
X.
2015
Modeling of soil water and salt dynamics and its effects on root water uptake in Heihe Arid Wetland, Gansu, China
.
Water
7
,
2382
2401
.
Liu
D.
2020
A rational performance criterion for hydrological model
.
Journal of Hydrology
590
,
1
15
.
125488. doi:10.1016/j.jhydrol.2020.125488
.
Liu
W.
,
Feng
Q.
,
Deo
R. C.
,
Yao
L.
&
Wei
W.
2020a
Experimental study on the rainfall-runoff responses of typical urban surfaces and two green infrastructures using scale-based models
.
Environmental Management
66
(
4
),
683
693
.
doi. 10.1007/s00267-020-01339-9
.
Liu
M. X.
,
Wang
Q. Y.
,
Guo
L.
,
Yi
J.
,
Lin
H.
,
Zhu
Q.
,
Fan
B.
&
Zhang
H. L.
2020b
Influence of canopy and topographic position on soil moisture response to rainfall in a hilly catchment of Three Gorges Reservoir Area, China
.
Journal of Geographical Sciences
30
(
6
),
949
968
.
doi:10.1007/s11442-020-1764-1
.
Livesley
S. J.
,
Dougherty
B. J.
,
Smith
A. J.
,
Luke
D.
,
Wylie
L. J.
&
Arndt
S. K.
2010
Soil-atmosphere exchange of carbon dioxide, methane and nitrous oxide in urban garden systems: impact of irrigation, fertiliser and mulch
.
Urban Ecosystems
13
(
3
),
273
293
.
Mohammad
V.
,
Sayed
M. B.
&
Jun
C.
2021
Global surface temperature: a new insight
.
Climate
9
(
5
).
No.81. doi:10.3390/cli9050081
.
Narjary
B.
,
Kumar
S.
,
Meena
M. D.
,
Kamra
S. K.
&
Sharma
D. K.
2020
Effects of shallow saline groundwater table depth and evaporative flux on soil salinity dynamics using Hydrus-1D
.
Agricultural Research
.
doi:10.1007/s40003-020-00484-1
.
Nie
W.
,
Ma
X.
&
Fei
L.
2017
Evaluation of infiltration models and variability of soil infiltration properties at multiple scales
.
Irrigation and Drainage
66
(
4
),
589
599
.
doi. 10.1002/ird.2126
.
Ribolzi
O.
,
Patin
J.
,
Bresson
L. M.
,
Latsachack
K. O.
,
Mouche
E.
,
Sengtaheuanghoung
O.
,
Silvera
N.
,
Thiébaux
J. P.
&
Valentin
C.
2011
Impact of slope gradient on soil surface features and infiltration on steep slopes in northern Laos
.
Geomorphology
127
,
53
63
.
doi:10.1016/j.geomorph.2010.12.004
.
Shi
D. M.
,
Wang
W. L.
,
Jiang
G. Y.
,
Peng
X. D.
,
Yu
Y. L.
,
Li
Y. X.
&
Ding
W. B.
2016
Effects of disturbed landforms on the soil water retention function during urbanization process in the Three Gorges Reservoir Region, China
.
Catena
144
,
84
93
.
doi: 10.1016/j.catena.2016.04.010
.
Song
Y. S.
,
Li
F.
,
Wang
X. K.
,
Xu
C. Q.
,
Zhang
J. Y.
,
Liu
X. S.
&
Zhang
H. X.
2015
The effects of urban impervious surfaces on eco-physiological characteristics of Ginkgobiloba: a case study from Beijing, China
.
Urban Forestry & Urban Greening
14
(
4
),
1102
1109
.
doi: 10.1016/j.ufug.2015.10.008
.
Sutanto
S. J.
,
Wenninger
J.
,
Coenders-Gerrits
A. M. J.
&
Uhlenbrook
S.
2012
Partitioning of evaporation into transpiration, soil evaporation and interception: a comparison between isotope measurements and a HYDRUS-1D model
.
Hydrology & Earth System Sciences
16
,
2505
2616
.
doi:10.5194/hess-16-2605-2012
.
Wiekenkamp
I.
,
Huisman
J. A.
,
Bogena
H. R.
,
Lin
H. S.
&
Vereecken
H.
2016
Spatial and temporal occurrence of preferential flow in a forested headwater catchment
.
Journal of Hydrology
534
,
139
149
.
doi:10.1016/j.j hydrol.2015.12.050
.
Wu
C.
,
Hao
Z. C.
,
Liu
C. Y.
&
Liang
J. P.
2018
Dynamic change regulation and simulation of soil moisture in Wudaogou Region
.
Water Resources and Power
36
(
9
),
138
142
.
(in Chinese).
Xiong
L.
,
Xu
Z. F.
,
Wu
F. Z.
,
Yang
W. Q.
,
Yin
R.
,
Li
Z. P.
,
Tang
S. S.
&
Xiong
H. T.
2014
Soil respiration of two typical urban lawns in Chengdu City during the winter dormancy period
.
Chinese Journal of Applied and Environmental Biology
20
(
2
),
275
280
.
(in Chinese)
Xu
Z. X.
&
Cheng
T.
2019
Basic theory for urban water management and sponge city-review on urban hydrology
.
Journal of Hydraulic Engineering
50
(
1
),
53
61
.
(in Chinese).
Yang
Q. H.
,
Chen
L. H.
,
Zhang
F.
&
Zhang
C. B.
2008
Responses of soil moisture variations to rainfall and vegetation
.
Journal of Beijing Forestry University
30
(
S2
),
88
94
.
(in Chinese).
Yang
S. M.
,
Zhang
T.
,
Zhao
Q. M.
,
Gao
X. Y.
,
Wang
Z. W.
&
He
T. B.
2020
Factors influencing ecosystem respiration in different cultivated grassland ecosystems in Guiyang
.
Pratacultural Science
37
(
11
),
2211
2222
.
(in Chinese).
Zang
W. B.
,
Liu
S.
,
Huang
S. F.
,
Li
J. R.
,
Fu
Y. C.
,
Sun
Y. Y.
&
Zheng
J. W.
2019
Impact of urbanization on hydrological processes under different precipitation scenarios
.
Natural Hazards
99
(
3-SI
),
12
1257
.
doi:10.1007/s11069-018-3534-2
.
Zhang
F. M.
&
Shen
S. H.
2007
Spatial distribution and temporal trend of reference crop evapotranspiration in China
.
Journal of Nanjing Institute of Meteorology
30
(
5
),
705
709
.
(in Chinese).
Zhang
R. X.
,
Zhao
X. Y.
,
Zhang
C. C.
&
Li
J.
2020
Impact of rapid and intensive land use/land cover change on soil properties in arid regions: a case study of Lanzhou New Area, China
.
Sustainability
12
(
21
).
No. 9226. doi:10.3390/su12219226
.
Zhao
Y. F.
,
Zou
X. Q.
,
Cao
L. G.
,
Yao
Y. L.
&
Fu
G. H.
2018
Spatiotemporal variations of potential evapotranspiration and aridity index in relation to influencing factors over Southwest China during 1960–2013
.
Theoretical and Applied Climatology
133
,
711
726
.
doi:10.1007/s00704-017-2216-4
.
Zhou
Q.
,
Huang
J. B.
,
Zhou
Y. M.
&
Huang
Y. Z.
2019
Variation characteristics of evapotranspiration and soil moisture in urban grassland: a case study on the regional grassland vegetation in Yangzhou City
.
Water Saving Irrigation
2019
(
3
),
22
26
.
(in Chinese).
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).