The sustainable use of groundwater resources concerning further climate change scenarios in Ulaanbaatar City Area , Mongolia

To estimate groundwater resources under changing climate is one of the important issues for Ulaanbaatar City in the Tuul river basin of Mongolia. The main water supply is provided from groundwater and demand has been increasing due to the rapid growth of population and economic development. There have not been any complete studies to assess climate change impact on groundwater resources for Ulaanbaatar city. Therefore, in this study we proposed to estimate future potential resources of the groundwater from the main wellfields in the city using the AnAqSim (Analytic Aquifer Simulator) model. The model calibration was performed on 10 wellfields during the reference period from 1960 to 2015. Based on the reliable calibration results for the natural conditions, the impact of climate change on groundwater resources was assessed to use the projected HadCM3 scenario for the periods 2046–2065 and 2080–2099. The results of the study contribute to a water management plan for the city to recommend seasonal abstraction.


GRAPHICAL ABSTRACT INTRODUCTION
Sustainable water supply is a crucial aspect of human life (Kristvik et al. ). Particularly, the regions where water supply is provided by groundwater will become more important due to climate change (Liuzzo et  In Mongolia, the mean annual air temperature has increased by 2.3 C since 1940 (MARCC ). The highest increase in air temperature was observed in winter periods at 3.6 C (Batima et al. ). Not only local changes but also regional climate models showed that the highest changes occurred in the Central and Eastern parts of Mongolia (Gomboluudev & Natsagdorj ). For precipitation change, the trend is heterogeneous but decreasing trends predominated in many regions (Dulamsuren et al. ; Natsagdorj & Gomboluudev ). Especially in July, the reduction of precipitation and increase in evaporation may affect the surface and groundwater regime (Sato et al. ). Once there is evidence of changing climate, groundwater responses need to be assessed in the country to support a better plan for groundwater management in the long term (Green et al. ). Specifically, Ulaanbaatar is the capital city of Mongolia which is located in central Asia. Administratively, the city consists of nine districts and covers an area of 4,700 km 2 . The population of Ulaanbaatar city has increased by 55% over the past 15 years (Enkhbat et al. ) and it is estimated that 1.4 million people now reside there. In the last 50 years, groundwater for Ulaanbaatar has decreased because of the rapid increase in water consumption as well as water scarcity caused by climate change, environmental degradation, and rapid urbanization (Dalai et al. ).
The groundwater system is an invisible phenomenon and there are several ways to assess climate change impact, including the numerical modeling approach (Dragoni & Sukhija  in Mongolia, two computer-based models were applied for estimation, such as Modflow based on the finite difference method (FDM) and unit specific yield (Buyankhishig et al. ), and AnAqSim based analytic element method (AEM) and groundwater pumping test parameters (Yihdego & Paffard ). AnAqSim (analytic aquifer simulator) is an analytic element software that simulates groundwater flow.
It uses subdomains described in Fitts (), which gives it strong capabilities concerning heterogeneity and anisotropy.
However, it uses some approximation for flow between aquifers and applies a radial basis function to estimate groundwater storage (Bakker ). Some studies suggested AnAqSim as a reasonable tool to model groundwater flow

Study area
We included 10 wellfields in the Ulaanbaatar city territory along the Tuul River valley (Figure 1). The Ulaanbaatar   For estimation of the groundwater resources, the following formula was used based on fundamental hydrogeological elements: where F is the area of the aquifer, m 2 ; H is depth of the aquifer, m; and n is porosity.
where s is drawdown; Q is pumping rate; T t is transmissitivy; r is the radial distance from the pumping well; and S is the storage coefficient.

RESULTS AND DISCUSSION
The groundwater head at every well was obtained from the model results. The contour map of the wellfields was made using simulated groundwater heads (Figure 3).
Also resulting from the 'steady state mode' of the model, the head boundary has been further refined to increase the accuracy of the modeled heads to within ±95% of the head difference.
The simulated groundwater heads were compared to the field observed heads at specific well locations (Figure 4).
Based on this, the average difference from the modeled head and observed data was around -2.7 m (range head 40 m). Figure 4 shows the plot between the observed and computed aquifer depth of the wells.
Moreover, aquifer area and depth of the wells were calibrated and values of the aquifer properties for the wellfields are illustrated in Table 3.
Using the results of the aquifer properties, the groundwater resources in the wellfields were calculated using the   further studies to focus the recharge estimation from the precipitation for the wellfields. Generally, AnAqSim gives an adequate understanding of the hydrogeological properties of the system using the observational data.
This study can be very helpful for groundwater professionals, policymakers, and city planners in deciding water security and sustainable groundwater use issues of the capital city of Ulaanbaatar, Mongolia.

CONCLUSIONS
The climate change impact on groundwater resources in the seven wellfields for drinking water supply and three wellfields for industrial use in the Upper Tuul basin over the Ulaanbaatar city was assessed using the AnAqSim modeling approach. The aquifer properties and potential groundwater resources of the wellfields were estimated. For this purpose, the area, average depth, and storage of the aquifer in the wellfields were calibrated based on the observational data.
The results show that the difference between the observed head and the simulated head is found to be in the 90% confidence level.
The average annual air temperature and precipitation were projected by HadCM3 with an increase of 1-5 C and 2-15%, respectively, by the end of the 21st century. Potential resources will be decreased by 0.3-10.3 and 1.2-14.8% for 2050 and 2080, respectively. In particular, the wellfields of the upper and downstream side of the river will be more sensitive to climate change. This means we should consider  proper groundwater use plans in the future to maintain sustainable use and for the ecosystem's wellbeing.

DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.