Impacts of Water Consumption in the Haihe Plain on the Climate of the Taihang Mountains, North China

Institute of Oceanographic Instrumentation, Qilu University of Technology (Shandong Academy of Sciences), Qingdao 266001, China Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing 100101, China Beijing Normal University, Beijing 100875, China Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China School of Geography and Tourism, Shaanxi Normal University, Xi’an 710119, China


Introduction
e Taihang Mountains, which extend over 400 km from north to south in North China, form a geographic boundary between the Loess Plateau (to the west) and the North China Plain (to the east).Mountains and intermontane valleys lie to the west and north, and plains lie to the east and south (Figure 1).A number of reservoirs in the Taihang Mountains are important water sources for North China.However, due to long-term human exploitation and climatic changes, the Taihang Mountains suffer from soil erosion and vegetation degradation [1,2].
During the past fifty years, significant drying and warming trends have been detected in North China [3,4].
e mean precipitation in the Taihang Mountains was found to decrease by 10.2 mm per decade, which is slightly less than the trend observed on the piedmont plain of these mountains, of 12.5 mm per decade [5].
Numerous studies have investigated climatic changes in North China and have mainly attributed these changes to human activity [6][7][8][9].For example, Jia et al. attributed the observed changes in water resources from 1961 to 2000 to various factors and demonstrated that local human activity accounted for about 60% of the observed changes [10].
For the Taihang Mountains, the most intense human activities occur in its eastern plains area, Haihe Plain, which is in the northern part of the North China Plain.e Haihe Plain is an important agricultural area in China and supports rapidly developing industries.According to the Urban Statistical Yearbook of China 2000, the population density and gross domestic product (GDP) in 2000 in the Taihang Mountains were 322 people per km and 239 × 10 4 yuan•km −2 , respectively.
is compares with 687 people per km and 530 × 10 4 yuan•km −2 in the Haihe Plain [11].e dense population and industry in the Haihe Plain have induced a serious imbalance between water demands and surface water supplies [12][13][14].Large amounts of groundwater have been exploited for crop cultivation and industrial development, leading to a rapidly dropping groundwater table and a series of eco-environmental crises [15][16][17].
Water consumption activities have been shown to enhance evapotranspiration and reduce local temperature, and moreover, increasing local water vapor from land may lead to more convection and further changes in atmospheric circulation [18][19][20][21][22]. Numerous studies have investigated the regional impacts of agricultural irrigation, which constitutes a major proportion of water consumption.For example, Saeed et al. applied the Max Planck Institute's Regional Model (REMO) and found that the irrigation over India caused increased evapotranspiration and less westerlies entering into land from the Arabian Sea [23].DeAngelis et al. analyzed the precipitation observations in the Great Plains of the United States and found the increased evapotranspiration due to irrigation contributed to more downwind precipitation [24].
For the Haihe Plain, Chen and Xie investigated the climatic effects of large-scale irrigation due to interbasin water transfer and demonstrated that such irrigation causes increased local precipitation and decreased temperature at the land surface [25].Leng et al. revealed that the changes in topsoil moisture and subsurface water flux induced by groundwater irrigation exceeded the potential change that could be attributed to various climate projections for the Haihe Plain [26].ese studies have made valuable progress in investigating the relationship between local climate changes and water consumption activities.However, the extent of the effect of water consumption within the Haihe Plain on its western water source area, the Taihang Mountains, remains unclear.Further discussion should address the way in which high-level water consumption in the eastern Haihe Plain affects the climate of the western Taihang Mountains.
erefore, in this study, a series of water exploitation and consumption simulations were conducted by the regional climate model RegCM4, using data from 1996 to 2007.
rough the comparison of these simulations, we investigated the effect of water consumption processes in the Haihe Plain on the spatiotemporal variability of the local climate of the Taihang Mountains.

Model Description.
e regional climate model used in this study, RegCM4, was developed by the International Center for eoretical Physics (ICTP) in Italy [27].RegCM4 is a hydrostatic climate model with a sigma vertical coordinate.ree convective precipitation schemes, Kuo, Grell and Emanuel, and a large-scale precipitation scheme, SUBEX, were available as simulation options.
e land surface models BATS (biosphere-atmosphere transfer scheme) and CLM (community land model) were available for use as land modules in RegCM4.
In this study, version 3.5 of CLM was used.e CLM model divides grid cells into multiple land units (glacier, wetland, vegetation, etc.).Vegetation units can be further divided into 17 plant function types (PFTs) [28].e hydrological processes in CLM3.5 were obtained from a runoff parameterization scheme with a simple groundwater model developed by Niu et al. [29,30].
e RegCM4 model has been implemented around the globe and demonstrated excellent climate simulation abilities.For example, Wang et al. evaluated the monthly precipitation simulation of RegCM4 over the Tibetan Plateau using the station observation and TRMM (Tropical Rainfall Measurement Mission) data and found RegCM4/ CLM3.5 performed better than RegCM4/BATS in the statistical indices [31].Ashfaq et al. used RegCM4 to dynamically downscale the historical and near-term future outputs from 11 global climate models, in order to present high-resolution ensemble projections of climatic changes over the continental United States [32].Mbienda et al. compared the simulations performed by different convective schemes of RegCM4 over Central Africa and found the Emanuel-MIT convective scheme showed better indices in temperature and the Grell scheme with Arakawa-Schubert closure assumption was better to downscale precipitation and surface wind [33].

Water Exploitation Scheme Description.
e scheme of water exploitation and consumption used in this study was developed based on the CLM3.5 approach used in our previous studies [34,35].
e scheme relies on preset water demand and is composed of exploitation and consumption components.As shown in Figure 2, the total water demand D t is supplied by water resources pumped from rivers (Q s ) and wells (Q g ).e total water demand D t is consumed by three consumption sections-irrigation consumption (D a ), domestic consumption (D d ), and industrial consumption (D i ).

2
Advances in Meteorology e irrigation water is treated as effective rainfall reaching the topsoil, as per the approach in previous studies [23,25,36,37].Furthermore, according to the descriptions of urban water use by Shiklomanov and the approach of specific hydrological modeling studies, the water for domestic and industrial consumption is highly simplified in the scheme to partly increase local evaporation and partly return to river channels [36][37][38].
e scheme is based on the balance of water demand and supply, which can be described as e surface water supply Q s is composed of the total runoff R t and stream discharge R str in each model grid cell, and it has a higher priority than the groundwater supply Q g for satisfying demand D t .When the surface water supply cannot meet the total demand, the groundwater supply Q g is subtracted from the groundwater storage and can be calculated as In the consumption section, the effective rainfall in the model increases by D a due to irrigation, which will further participate in the processes of runoff generation and infiltration.e increased evaporation in the model is defined as ε(D d + D i ) and the remaining water of domestic and industrial consumption, (1 − ε)(D d + D i ), is regarded as wastewater recharge returning to river channels.
In this study, domestic and industrial water consumption occurred year-round, while irrigation consumption only occurred during the main crop growing periods.In North China, the main crops are winter wheat and summer maize.According to their known phenology, irrigation first occurred from March 10 (when wheat turns green) to June 18 (wheat harvest time).Irrigation also occurred from June 22 to August 28 (summer maize planting and harvesting times, respectively) [39][40][41][42].
e evaporation rate of domestic and industrial consumption ε was set as 0.26 based on the statistical value of water consumption rate for industrial and urban domestic consumption in the Water Resource Bulletin of China for the year 2000 (http://www.mwr.gov.cn/sj/tjgb/szygb/201612/t20161222_776035.html).

Water Demand Estimation.
Due to the lack of spatial distribution data, the water demand in each grid cell was indirectly estimated from relevant eco-social data, as other studies have done [36].In this study, the total demand D t was divided into three components-domestic, industrial, and irrigation demand-which were estimated based on data of population, GDP, and irrigated cropland areas in China for the year 2000.e estimation equation can be expressed as where D t,2000 is the total water demand in the year 2000 (kg•m −2 •s −1 ); ρ is water density (1 × 10 3 kg•m −3 ); λ is the number of seconds in a year (31,536,000 s•yr −1 ); S is the grid cell area (m 2 ); c 1 is per capita annual domestic water consumption (m 3 •person −1 •yr −1 ); A pop is the population in each grid cell (persons); β is the conversion ratio between GDP and industrial output; A GDP is GDP in each grid cell (yuan); c 2 is industrial water consumption for unit industrial output (m 3 •yuan −1 •yr −1 ); α is the ratio between the irrigated and total cropland areas; c 3 is annual irrigation water consumption per hectare (m 3 •ha −1 •yr −1 ); and A agr is the cropland area in each grid cell (ha).Population and GDP data for the year 2000 were obtained from the Data Center for Resources and Environment Advances in Meteorology of Sciences, Chinese Academy of Sciences (http://www.resdc.cn).Cropland area data for 2000 were obtained from the China land cover classification dataset available at the Science Data Center for Cold and Arid Regions (http://westdc.westgis.ac.cn) [43].e three eco-social datasets had a resolution of 1 km × 1 km and were then interpolated to suit the RegCM4 model's resolution of 20 km × 20 km for the subsequent simulation tests.e parameters of water consumption were collected from the Water Resource Bulletin of China for the year 2000, where the provincial mean values were available.
e other two parameters, β and α, were collected from the China Statistical Yearbook 2000 and were also provincial mean values [44].
e spatial distribution of estimated water demand D t,2000 per unit area is shown in Figure 3(a), and the demands for the Taihang Mountains (Figure 3(b)) and Haihe Plain (Figure 3(c)) were then cropped from Figure 3(a).As shown in Figure 3(b), more water is consumed in the eastern piedmont areas and south-central valleys of the Taihang Mountains.For the Haihe Plain (Figure 3(c)), and areas near Beijing (116.5 °E, 40.0 °N) and Tianjin (117.0 °E, 39.0 °N), there were higher water demands due to intense industrial activity and high population densities.e central areas of the plain are the main agricultural areas, and water demands there are also high.
Table 1 provides a comparison of the estimated and actual total water demands in 2000 for China and the Haihe River Basin.e estimation and actual values are basically consistent with each other, although they do not strictly coincide due to the use of different data sources.For the whole of China, the estimated irrigation demand provides most of the error, because of inconsistencies in cropland area between the remote sensing data and China Statistical Yearbook data used in this study.In addition, a comparison for the Haihe River Basin is also presented, because the whole Haihe Plain and most areas of the Taihang Mountains lie in the basin, and few data for the Taihang Mountains are available in water resource bulletins.
To obtain the demands during the simulation period (1996-2007), the mean annual water use for the Haihe River Basin was derived from the Water Resource Bulletin of Haihe River Basin (http://www.hwcc.gov.cn/hwcc/wwgj/xxgb/). Using these statistics, the estimated water demands in 2000 for the Taihang Mountains and Haihe Plain were scaled up or down, keeping the three water-consuming sectors unchanged, to approximate interannual variations.e estimated annual demands from 1996 to 2007 are shown in Table 2.

Experimental Design.
e simulation domain of RegCM4 is shown in Figure 1, where the domains of Taihang Mountains and Haihe Plain are outlined by black lines.e central projection of RegCM4 was located at 116 °E, 38 °N, and it used a spatial resolution of 20 × 20 km. e Grell scheme was used as the convective precipitation scheme, and ERA-Interim reanalysis data from 1992 to 2007 were used as the lateral boundary forcing.e time steps were 30 seconds for the atmosphere module and 30 minutes for the land surface module of CLM3.5.Before the simulations were conducted, a test simulation using the original RegCM4 with ERA40 reanalysis data from 1961 to 1991 was conducted to obtain the balanced groundwater depth.A spin-up simulation from 1992 to 1995 was then conducted using the original RegCM4 without exploitation.Based on the final status of the spin-up simulation, four simulation tests were conducted from 1996 to 2007.A control test (CTL) continued to simulate the natural state without exploitation; Exploitation Test 1 (T1) used estimated water demand data for the Haihe Plain, and Exploitation Test 2 (T2) used double this demand.In addition, Exploitation Test 3 (T3) used combined demand data of the Taihang Mountains and Haihe Plain.By comparing differences between the exploitation and control tests, this study aimed to investigate the effects of water consumption activities in the Haihe Plain on local climate changes in the Taihang Mountains.

Validation of the Control Test.
Station observed precipitation, and 2 m air temperature data during the simulation period was derived from the China National Meteorological Information Center (http://data.cma.cn/en) and then was interpolated as gridded data with a resolution of 0.5 °× 0.5 °, by using the inverse distance square weighting method as preformed in previous studies [45,46].e spatial distributions of annual mean precipitation and 2 m air temperature from 1996 to 2007 are shown in Figure 4. e CTL test run in RegCM4 simulated the regional climatology over the study domain well.Specifically, the simulated precipitation and temperature increased from northwest to southeast, which is consistent with observations.However, slight low biases were detected in the Taihang Mountains, of −25.1 mm/year for precipitation and −0.64 K for 2 m air temperature.Also, a dry and warm bias, of −115.3 mm/year for precipitation and 0.88 K for 2 m air temperature, was detected in the Haihe Plain.Additionally, the statistical indices of monthly series in the Taihang Mountains and Haihe Plain are listed in Table 3, including the temporal correlation coefficient, root-mean-square error, and standard deviation.ese statistics indicate that the simulated series corresponds with observations; hence, the RegCM4 model is suitable for simulating temporal variations in the local climate.in the eastern piedmont areas and the south-central valleys, corresponding with the higher water demand there, as shown in Figure 3(b).e processes of water exploitation and consumption withdraw groundwater and consume it at the surface, which leads to increased upper soil moisture and humidity of the lower atmosphere.Increased moisture at the land surface may further enhance regional wetting effects via vertical convective activity and horizontal transport in the upper troposphere.

Spatial Distribution of Climatic Differences
e differences in soil moisture at 10 cm depth between the exploitation and control tests are shown in Local wetting changes at the land surface in the Haihe Plain led to changes in precipitation, not only in the local plain areas, but also in the surrounding areas (including the Taihang Mountains) via atmospheric moisture transfer.
As shown in Figure 5(g), increased precipitation was detected in most areas of the Haihe Plain, with a mean value of 0.046 mm/d.e mean precipitation also increased by 0.020 mm/d in the Taihang Mountains, and the spatial distribution of increased precipitation was in accordance with the distribution of soil moisture (Figure 5(d)).
e distribution of precipitation differences in Figure 5(h) is similar to that in Figure 5(g) but with approximately double values, specifically, 0.030 mm/d in the Taihang Mountains and 0.089 mm/d in the Haihe Plain.For the T3 test (Figure 5(i)), when the water consumption processes in the Taihang Mountains were considered in conjunction with the processes in the Haihe Plain, the precipitation in the Haihe Plain was slightly enhanced to 0.050 mm/day due to more land water over a larger scale being transferred into the atmosphere.In addition, precipitation increased by 0.028 mm/day in the Taihang Mountains, which is slightly higher than the mean difference in the T1 test, due to local increased evaporation.
Accompanying the regional wetting effects, the water consumption processes also led to cooling effects via changes in the heat fluxes emitted from the land surface.As shown in Figure 6(a), the mean 2 m air temperature decreased by 0.613 K in the Haihe Plain due to the local water consumption of the T1 test.Weak cooling effects were also detected in the Taihang Mountains, with a mean temperature decrease of 0.049 K.For the T2 test with doubled demand (Figure 6(b)), the changes in temperature were slightly less than double the changes observed in the T1 test.ere were decreases of 1.121 K in the Haihe Plain and 0.089 K in the Taihang Mountains.For the T3 test (Figure 6(c)), the cooling effects were slightly enhanced in the Haihe Plain.
e mean temperature decrease was 0.646 K in the Haihe Plain and 0.201 K in the Taihang Mountains.
In CLM, changes in the upward heat fluxes emitted from the land surface are reliant on differences in temperature and humidity between the land surface and the lower atmosphere.
us, locally wetter and cooler air, due to water consumption in the Haihe Plain, was transferred to the Taihang Mountains via atmospheric activity.Subsequently, changes in land-atmosphere differences in temperature and humidity further affected the heat fluxes in the Taihang Mountains.erefore, the changes in the heat fluxes of the Taihang Mountains, which were indirectly affected by the wetter and cooler atmosphere of the Haihe Plain, were far less than the changes in the Haihe Plain.For the T1 test (Figure 6(d)), the sensible heat flux decreased by 7.682 W/m 2 , on average, in the Haihe Plain, yet only decreased by 0.098 W/m 2 in the Taihang Mountains.For the T2 test (Figure 6(e)), the mean decrease was 13.884 W/m 2 in the Haihe Plain and 0.168 W/m 2 in the Taihang Mountains.If local water consumption is considered, the heat fluxes will be affected significantly due to the increased evaporation associated with water consumption processes.us, the sensible heat flux in the T3 test (Figure 6   , respectively.e sum of sensible and latent heat flux was positive in most areas of the Taihang Mountains, indicating that the total heat flux emitted into the atmosphere from the land surface increased and the temperature of the soil layers would subsequently change.

Vertical Profiles in Soil and
Atmosphere Layers.To investigate changes in vertical profiles, Figure 7 provides profiles of mean soil moisture and temperature differences in the Taihang Mountains.e profiles of soil moisture differences (Figure 7

􏽱
), x mi and x oi are the simulated and observed monthly mean precipitation or 2 m air temperature in the month i, and n is the number of months; RMSE: root-mean-square error, RMSE Advances in Meteorology reaching the topsoil layer.e differences between the T1 and T2 tests in each soil layer basically remain constant.For the T3 test, which considered local water exploitation and consumption, soil moisture increased significantly in the upper layers due to increased water fluxes (e.g., irrigation and precipitation) into the top layer.In CLM, drainage from soil layers to aquifers is reliant on a water potential gradient between the soil layers and aquifers.is drainage will gradually approach the maximum infiltration capacity with increasing groundwater table depth.erefore, in the lower soil layers, soil moisture was detected to decrease with depth.
Unlike the characterization of soil moisture, the soil bottom is regarded as being heat-insulated in CLM, and soil temperature is only dependent on the energy balance at the surface.As shown in Figure 7(b), soil temperature differences basically remained constant with depth in the T1 and T2 tests, and the differences in the T2 test were slightly less than double the differences in the T1 test.For the T3 test, due to the continuous water consumption process, the temperature decrease in the upper soil layers was greater than that in the lower layers, and the decrease tends to be constant with depth.
While domain-averaged profile curves are provided in Figure 7, latitude-averaged contours of atmospheric temperature and humidity between 35 °N and 40 °N are provided in Figure 8 to investigate the transfer of cooling and wetting effects into the atmosphere.As shown in Figure 8(a), increased air humidity over the Haihe Plain (approximately 114.5 °E−119 °E) below a pressure of 800 hPa was detected, and at a pressure of around 600 hPa, a center of dry values was detected because the convergent flow from the lower atmosphere becomes divergent at 600 hPa.For the T2 test (Figure 8(b)), the wetting effects in the lower atmosphere were enhanced, and more moisture was transferred to the atmosphere over the Taihang Mountains (approximately 111 °E−114.5 °E).
e wetting effects in the south-central valleys of the Taihang Mountains (about 112.5 °E) were even stronger in the T3 test (Figure 8(c  8 Advances in Meteorology effects in the Haihe Plain.When compared with the changes in the T1 test (Figure 8(a)), the atmospheric wetting effects were enhanced over the Haihe Plain, and this slight enhancement was in accordance with the mean differences in climatic variables shown in Figures 5 and 6. e profiles of air temperature differences (Figures 8(d)-8(f)) were similar to the air humidity profiles; that is, water consumption processes in the Haihe Plain lead to regional cooling effects in the atmosphere; not only over the Haihe Plain, but also over its surrounding areas, including the western Taihang Mountains.e cooling effects were basically below the 800 hPa level and were enhanced as the water volume consumed at the surface increased.Also, the cooling effects over the Haihe Plain were enhanced slightly due to water consumption over a larger-scale domain.

Annual and Seasonal Variability.
Figure 9 shows the mean annual series of differences for the Taihang Mountains.e series in Figure 9(a) indicates that the increased groundwater depths in the T1 and T2 tests were almost negligible and the groundwater depth dropped continuously due to local groundwater exploitation in the T3 test.For the other series of variables, the differences between the T1 and T2 tests indicate that doubled water consumption in the Haihe Plain did not cause doubled changes in the Taihang Mountains in most years.A negative feedback mechanism exists between increased water demand and subsequent climatic changes.
Meanwhile, the differences between the T1 and T3 tests indicate that changes in land surface variables due to water consumption are usually far greater when the water is consumed locally rather than in neighboring areas.e differences between the T1 and T3 tests also differed between variables.For the atmospheric variables, these differences were the lowest, because cooling and wetting effects occurring in the Haihe Plain can be transferred to neighboring areas via atmospheric activity, including to the Taihang (W/m 2 ) (W/m 2 ) (W/m 2 )   Advances in Meteorology Mountains.For example, the precipitation data series of the T1 test (Figure 9(c)) was approximately 35% lower than the T3 series.e differences between the T1 and T3 tests were higher for the land surface variables, which were further affected by the cooler and wetter atmosphere.For example, the T1 series of 10 cm soil moisture (Figure 9 greater during the irrigation seasons (primarily spring and summer) due to increased precipitation, and during the nonirrigation seasons (primarily autumn and winter), the increased soil moisture reduced gradually due to the waterholding capacity of soil.When considering local water consumption processes, 10 cm soil moisture was detected to increase clearly due to local irrigation.
For the other four variables, most differences occurred during the irrigation seasons, regardless of whether local water consumption processes were considered.During  12 Advances in Meteorology nonirrigation seasons, the wetter and cooler differences still existed in the T3 test, because increased evaporation due to domestic and industrial consumption can also lead to wetter and cooler differences to a certain level.However, the differences in the T1 and T2 tests were almost negligible during the nonirrigation seasons because the cooling and wetting differences were weak, even in the Haihe Plain, and low temperatures in nonirrigation seasons also suppress convection and moisture transfer from the Haihe Plain to the neighboring Taihang Mountains.

Conclusion and Discussion
In this study, the regional climate model RegCM4 was incorporated with a scheme of human- (1) e processes of water exploitation and consumption in the Haihe Plain cause rapid increase of groundwater depth and local cooling and wetting effects at the land surface. is causes a cooler and wetter atmosphere that transfers to surrounding areas, including the Taihang Mountains, leading to decreases in temperature and increases in surface moisture.e cooling and wetting changes at the land surface of the Taihang Mountains affect the local energy balance, increasing latent heat flux, and decreasing sensible heat flux emitted from the land surface.(2) ese wetting and cooling effects are positively related to the volume of water consumed.However, the cooling effects do not enhance the development of lower atmospheric convection and moisture transfer.For the T2 test with double the water demands of the T1 test, the wetting and cooling changes were slightly suppressed and were less than double the changes estimated in the T1 test.
(3) In the simulations, water consumption activities in the Haihe Plain-induced wetting and cooling effects in the Taihang Mountains via atmospheric transfer.However, these wetting and cooling effects were rather weak at the land surface and caused less change than that caused by local water exploitation and consumption in the Taihang Mountains.For the T1 and T2 tests, changes at the land surface in the Taihang Mountains were entirely caused by changes in atmospheric variables.For comparison, the conditions of the T3 test were the opposite-local water consumption activities first led to changes of surface variables, then atmospheric variables changed due to land-atmosphere interactions.(4) e cooling and wetting effects of water consumption activities mainly occur during irrigation periods, because irrigation comprises most of the total water demand in the Haihe Plain and Taihang Mountains.
is study simulated the impacts of water consumption activities in the Haihe Plain on the local climate in the Taihang Mountains, located immediately to the west.However, many aspects of the tests, including data estimation, scheme design, and structural defects in RegCM4/CLM3.5,introduced uncertainties to the simulations and are discussed next.
Although the resolution of the water demand data used in this study was higher than the one used in our previous studies, there was potential underestimation of the total demand according to Table 1.
e largest uncertainty remains the estimation of irrigation water use.In this study, the ratio between the irrigated and total cropland areas, α, was constant derived from the China Statistical Yearbook 2000.Future work should address the acquisition of more reasonable distributions of α by using remote sensing data.e Water Resource Bulletin of China further indicates that the per capita water use differs greatly between urban and rural regions.e lack of consideration about the rural-urban difference may lead to excessive domestic water demand estimation over rural regions; thus, a further division between rural and urban domestic demand should be addressed to differentiate the allocation proportion of domestic water consumption.e scheme of water exploitation and consumption used in this study was highly simplified, and more detailed data should be obtained to improve the spatial heterogeneity of water demand estimation and the allocation process of water consumption.
Structural defects in the RegCM4/CLM3.5 models also introduced uncertainties.For example, water in lakes and reservoirs was not included in the water cycle process of the model.Although significant improvements were made by introducing the lake depth dataset, more accurate lake parameterization, and groundwater recharge processes, the water volume in lakes and reservoirs was still maintained constant even in the newer versions of the CLM model [47][48][49].us, the surface water supply was only withdrawn from river channels, leading to overexploited groundwater and deeper groundwater table in the simulation tests.A mechanism of reservoir regulation based on the accurate description of the reservoir water volume should be included in the river transport model component of CLM. e processes of groundwater lateral flow were also not included in the model, and the groundwater table depth was only dependent on water input/output in the vertical direction.
e lack of groundwater lateral flow data input may lead to groundwater grid cells being isolated from each other and to spatial distributions of simulated groundwater depths differing from actual conditions, especially in mountainous areas.Future efforts should be made to incorporate 3-D groundwater flow into the CLM model.
Furthermore, the time frame of the study is relatively old.e year 2000 is used as the benchmark year and the Advances in Meteorology simulation tests are conducted for the years 1996-2007.As there are significant demographic changes in the areas under study during the past decade, it would definitely be of interest to explore the climatic effects of water exploitation and consumption using the updated numbers of population, GDP, and water use. is could be the focus of a future study; however, here we discuss a few preliminary aspects.In fact, the water use over China in general and the Haihe Plain in particular did not increase significantly from the year 2000 onwards.e rapid increase in water use that took place in the previous decades (mainly the 1980s and 1990s) was discontinued upon the beginning of the new millennium.According to the Water Resource Bulletin of China over the years, the total water volume used all over China was 556.6 billion m 3 in 1997 as opposed to 604 billion m 3 in 2016.
erefore, and despite the significant increase of the Chinese GDP after the year 2000, the total water use was characterized by a relatively small raise.And it is widely known from previous studies that the climatic impacts of water consumption rely heavily on the volume of the water consumed [18,19,23,25].Moreover, our study uses multiyear mean differences of the simulation tests to discuss the mechanism and magnitude of the climatic impacts of water consumption, which are not expected to change significantly if the climate forcing data of RegCM4 is replaced with a recent dataset.erefore, based on the aforementioned features, it would be expected that the results of our simulation tests would not change significantly if more recent data was to be used.

Figure 1 :
Figure 1: Topography of the simulation domain.e Taihang Mountains and Haihe Plain are outlined in black.

Figure 2 :
Figure 2: Framework of the water exploitation and consumption scheme.

Figure 3 :
Figure 3: Spatial distributions of estimated water demand per unit area in 2000 over (a) all of China, (b) the Taihang Mountains, and (c) Haihe Plain.
Figures 5(d)-5(f).e changes in soil moisture at 10 cm depth were highly dependent on changes in precipitation and irrigation water reaching the soil surface.As shown in Figure 5(d), the soil moisture increased by a mean of 0.015 m 3 /m 3 in the Haihe Plain, and a large increase in soil moisture appeared in the central areas, where irrigation demands occupy a high proportion of the total demand.Wetting effects in the Haihe Plain also led to an increase in soil moisture in the Taihang Mountains; however, this increase was only 0.002 m 3 /m 3 , because the increase was entirely from increased local precipitation.For the T2 test with doubled water demand (Figure 5(e)), soil moisture increased by 0.040 m 3 /m 3 in the Haihe Plain and 0.003 m 3 /m 3 in the Taihang Mountains.Although the water consumption processes in the Haihe Plain led to increased soil moisture in the Taihang Mountains, the extent was weaker than that of the wetting effects caused by local water consumption processes.Considering the local water consumption in the Taihang Mountains (Figure 5(f)), the soil moisture increased by 0.016 m 3 /m 3 in the Haihe Plain and 0.008 m 3 /m 3 in the Taihang Mountains.
(f )) decreased by 2.837 W/m 2 in the Taihang Mountains and by 7.793 W/m 2 in the Haihe Plain.e changes of latent heat flux shown in Figures 6(g)-6(i) were similar to the changes of sensible heat flux except for their larger magnitudes.In the Haihe Plain, the latent heat

Figure 4 :
Figure 4: Spatial distributions of mean (a) observed precipitation, (b) precipitation simulated by the control test, (c) observed 2 m air temperature, and (d) 2 m air temperature simulated by the control test.

Figure 7 :Figure 8 :
Figure 7: Mean profiles of (a) soil moisture difference and (b) soil temperature difference in the Taihang Mountains.

Figure 9 :
Figure9: Annual series of (a) groundwater depth differences, (b) soil moisture differences at 10 cm depth, (c) precipitation differences, (d) 2 m air temperature differences, (e) sensible heat flux differences, (f ) latent heat flux differences in the Taihang Mountains.

Figure 10 :
Figure 10: Mean monthly (a) soil moisture differences at 10 cm depth, (b) precipitation differences, (c) 2 m air temperature differences, (d) sensible heat flux differences, and (e) latent heat flux differences in the Taihang Mountains.

Table 2 :
Actual statistical water use over the Haihe River Basin and estimated water demands over the Taihang Mountains and Haihe Plain from 1996 to 2007 (unit: 10 8 m 3 /year). 2

Table 3 :
Statistics of control test in the Taihang Mountains and Haihe Plain.
)) than the wetting induced water exploitation and consumption to investigate the impact of water consumption activities in the Haihe Plain on the local climate in the Taihang Mountains.Four simulation tests (i.e., three exploitation tests and one control test) were conducted.One Exploitation Test (T1) considered water consumption activities in the Haihe Plain and another (T2) doubled these water demands.ethird Exploitation Test (T3) considered the combined water demands of the Haihe Plain and Taihang Mountains, and a control test (CTL) simulated a case without any water exploitation or consumption.By comparing the differences between these three exploitation tests and the control test, we were able to analyze changes to climatic variables in the Taihang Mountains due to water consumption activities in the neighboring Haihe Plain.e main conclusions are as follows.