Assessment of Water Balance at Mayang Watershed, East Java

Mayang Watersheds frequently hit by floods during the rainy season and drought during the dry season. This study aims to assess the water balance by calculating water resource availability and water demand in the Mayang watershed. The Water Evaluation and Planning (WEAP) model was used as the primary tool for the analysis. The supply of water comes only from precipitation. Demand was calculated based on the water demand for irrigation, domestic, urban, industrial, and livestock uses. The unit of time to calculate the water balance is ten days. It means that each month is divided into three-time steps. Analysis of the WEAP is based on the water demand from 2002 to 2019. The results showed that from 3rd December to 1st May, the Mayang river and its tributaries could supply all demand sites up to 100%. However, unmet demand occurs from 2nd May to 2nd December. The highest first unmet demand occurred in October, with 0.67 million m 3 . The management of water resources, especially in terms of distribution during the rainy season and dry season, must be considered. and Several affect activities WEAP can analyze the WEAP can analyze the and of The analysis results can be used as a basis for decision-making in watershed management for the future.


Introduction
Water is an essential requirement for human survival; without water, there would be no life on earth. Water demand is water that is used to fulfill daily activities. Water sources for daily needs, in general, must meet the quantity and quality standards (Kencanawati & Mustakim, 2017). Water resources on the watershed use to supply water demand for irrigation, domestic, urban, industrial, livestock, and other needs. Water becomes a valuable resource, both in terms of quality and quantity (Misra, 2014). A Watershed is an ecosystem that has elements including natural and human resources. Natural resources act as objects consisting of land, vegetation, and water, while the human element is the subject or actor of the utilization of the elements of natural resources (Zuriyani, 2017). Regional development will cause water demand to increase in line with population growth. Fulfilling food needs and population activities is always closely related to water demand. Thus, water resources Geosfera Indonesia 57 Ariska Mia Christiwarda Sihombing et al. / Geosfera Indonesia 6 (1), 2021, 55-76 that is modeled using the WEAP. Several researchers have focused on the WEAP application.
For example, WEAP calculates the water balance of the Rawatamtu Sub-watershed (Setiawan et al., 2019). This research is limited to agricultural water balance. Our study intends to calculate the overall water balance in the whole area of the Mayang Watershed. In this study, the water balance component consists of water demand from various fields, such as irrigation, domestic, urban, industrial, and livestock. Therefore, this study calculates the water balance in more detail. This study aims to calculate the availability of water resources and water demand in the Mayang watershed.

Study Site and Input Data
The Mayang watershed ( Figure 1) has an area of 1,135 km 2 . The altitude on the watershed varies from 95 to 3,175 masl (meter above sea level). More than 95% of the watershed area is located at Jember Regency, while only a few areas are located at Banyuwangi Regency. Both hydrological data series and geo-spatial are used as input for modeling in the WEAP. The hydrological data series consists of precipitation and water demands data.
Precipitation data were selected from seven (7) measurement sites having the most extended recording periods (i.e., 2002-2019). Figure 2 presents the location of the seven (7) precipitation stations on the watershed. Data series related to irrigation water demand is 58 Ariska Mia Christiwarda Sihombing et al. / Geosfera Indonesia 6 (1), 2021, 55-76 obtained from the local water agency. Furthermore, water demands from other sectors, i.e., domestic, urban, industrial, and livestock, are calculated from existing statistical data.
The Geo-spatial data includes the Digital elevation model (DEM), land cover map, and soil type map. The DEM is extracted from the DEMNAS (Badan Informasi Geospasial, 2019). The DEMNAS is the digital elevation model data at the national level that can be freely downloaded from the National Geospatial Information Agency (Badan Informasi Geospatial or BIG) through their official website (Badan Informasi Geospasial, 2019). The DEMNAS is used to delineate the watershed boundary, determine the river network, and describe the water balance schematic (supply vs. demand) on the WEAP platform. This map uses to describe the condition of the watershed for the periods from 2011 to 2020. Finally, the soil type layer was obtained from the existing database (Soil Research Institute, 1998) and represented the soil layer condition.

WEAP
The WEAP model is integrated modeling software that simulates and calculates water supply, water demand, and environmental requirements and considers the impact of water quantity, water quality, and ecosystem policies. WEAP considers supply preferences and demand priorities to solve water allocation problems using a combination of linear and 59 Ariska Mia Christiwarda Sihombing et al. / Geosfera Indonesia 6 (1), 2021, 55-76 heuristic programming (Sieber & Purkey, 2015). WEAP can be used as a tool to support decisions in water resource management (Le Page et al., 2012) 2.3 Procedure

Analysis of Precipitation Data
The consistency test was calculated for each precipitation station (Ilham et al., 2018).
This process needs to be done because the rainfall required to prepare a water use design is the average rainfall in all areas concerned, not rainfall at a certain point (Mangende et al., 2016). Table 1 present the test result for the seven stations. It can be concluded that the seven stations are consistent, which means that the measured and calculated data are correlated.
Then, point precipitation measurement is then interpolated to areal precipitation by using the Thiessen polygon interpolation method. The interpolation was prepared using the existing tutorial.  (Hakim, 2019).
The land cover data is used to simulate the hydrological and to calculate evapotranspiration. Table 2 shows the land cover class distribution (in km 2 ) and the change    The crop data used to determine the Palawija Relative Area (PRA). PRA is the ratio of water requirements between one crop type to another (Haliem et al., 2013). Then, the PRA is used to calculate the irrigation water demand. Furthermore, the coefficient for each crop type is determined by using Table 3.  Figure 4. The soil type map clip from the soil layer database as published by (Soil Research Institute, 1998). Furthermore, the PRF values for each type of soil class are presented in Table 4.  Finally, the demand for irrigation (Qir) is obtained from the multiplication of the PRA and PRF which can be written on eq. 2 (Dewi et al., 2014): where, Qir: irrigation water demand (l/s); PRA: palawija relative area (ha.pol); PRF: palawija relative factor (l/s/ha).

b. Domestic, Urban, and Industrial Water Demand
First, the domestic water demand (DWD) is needed by households. The households usually withdrawal water directly by a sink or borehole from the unsaturated zone.
Otherwise, households obtained water from the Municipal water supply system service (BSN, 2015). In this case, the DWD is calculated based on the number of residents in the area (Mashuri et al., 2015). The main factor determining domestic water demand is population growth (Afrianto et al., 2015). DWD depends on the city category based on the population in liters/person/day, as presented in Table 5 (BSN, 2015). The equation for calculating domestic water demand can be written in eq. (3): Second, the urban water demand (UWD) is defined as the water needed to support the social and commercial facilities, such as hospitals, hotels, schools, shops, and warehouses are assumed to be 15% -30% of the total DWD (BSN, 2015). The equation for calculating urban water demand can be written in eq. (4) : Thirdly, Industrial water demand (IWD) assumed tends to be constant with time. The more the industry increases, the more water demand increases (BSN, 2015). Data of industrial water demand (IWD) is obtained from the local public water service at the regency level during the field survey.
c. Livestock The livestock water demand (LWD) is calculated based on the number and types of livestock in the watershed area (Zulkipli et al., 2012). The LWD is determined by the number and growth rate of livestock (Putri et al., 2016). For this reason, an analysis is needed to estimate the number of livestock in the next fews years. The amount of LWD is presented in Table 6. The eq. (5) is used for calculating LWD : where, LWD = livestock water demand ( l /day), q1 = water requirements for types of livestock and horses (l/day), q2 = water requirements for goats and sheep (l/day), q3 = water requirements for poultry (/day) P1 , P2, P3, = number of types of livestock.

Preparing the WEAP scheme
The boundary of the watershed was derived from the DEMNAS using a hydrological function and WEAP tool. A blue line represents the river flow network. A point in red represents the location of the output points. A demand site relates to a set of water users in an area (Sieber & Purkey, 2015). The WEAP scheme model of the Mayang watershed is shown in Figure 5.

Calibration and Validation
The calibration processes produce a coefficient of determination (R 2 )= 0.78 ( Figure   6), while for validation result of R 2 = 0.88 (Figure 7). Both R 2 values relatively acceptable to measure the correlation between simulated and observed discharge. The average volume of surface runoff per period is 29.75 million m 3 . Figure 8 shows that the dry season can start from 1st May to 1st November, followed by the rainy season (monsoon), starting from November to 2nd April.  (Table 2). This causes water that falls as precipitation less infiltrate to the soil layer. The storage in the soil layer (groundwater) decreases, and therefore more precipitation is converted directly to runoff.
The amount of annual precipitation from 2002 to 2019 is illustrated in Figure 12.
The average annual precipitation is 1,733 mm. However, the maximum annual precipitation can reach 4,000 mm per year, while the minimum can go down to 632.23 mm. The average annual precipitation is 1,733 mm. The average annual evapotranspiration is 632.23 mm and for runoff is 920.61 million mm. The precipitation trend in the Mayang watershed area from 2002-2019 has increased, although slightly. This increase was also found by Setiawan & Hariyanto (2017) in several regions in East Java. It means that the amount of precipitation received is sufficiently available. However, the distribution of precipitation between rainy and dry seasons is significant. Therefore, the leading water resources management problem is how to deal with the inequality of water distribution between dry and rainy seasons. In other words, the problem of water resources management is more accentuated by water balance.

Irrigation Water Demand
The average volume of irrigation demand per 10-day period from 2002 to 2019 is presented in figure 10. It is noted that per the 10-days, the irrigation water demand is about 1.48 million m 3 .  Year 69 Ariska Mia Christiwarda Sihombing et al. / Geosfera Indonesia 6 (1), 2021, 55-76 statement was supported by Putri et al. (2016), who stated that another aspect that affects the irrigation water demand is climate.

Domestic, Urban, and Industry Water Demand
The average water demand volume (per 10-day interval) for domestic, urban, and industrial from 2002 to 2019 is about 0.94 million m 3 (Figure 12). The average annual volume of domestic, urban, and industrial water demand is 33.97 million m 3 (Figure 13). development, population density, and the increase of public facilities. This was also found by Afrianto et al. (2015), where the increase in population impacted increasing water demand.
The phenomena may continue for the next years as the population always grows.

Livestock
The average water demand volume (per 10-day period) to support livestock production from 2002 to 2019 is 0.11 million m 3 (Figure 14). The annual water demand for livestock production from 2002 to 2019 varies from 3.7 to 4.8 million m 3 , while the average annual demand is 4.07 million m 3 (Figure 15). The coverage is the percentage of each site demand met, from 0% (no water) to 100% (sufficient water supply). Coverage 100% means that the supply can cover all water demand. Figure 16. Coverage Figure 16 shows that from 2nd May to 2nd December, the supply is insufficient to meet the demand. It is supposed that 1st May is the start of the dry season, and therefore the lack of water supply starts on 2nd May. This happens because of the low precipitation due to the dry season. Apart from low precipitation, the need for water in 10 days is also high due to the three times rice planting pattern in a year in the study location. As a result, there is a severe shortage of water. On 1 st October, the peak was only 74.92% of water demand can be Year The water demand for each site tends to increase every year (i.e., domestic, urban, and industrial water demand sites). The increase of population on the watershed is expected to be quite large, considering an increase of 0.55% (Badan Pusat Statistik Kabupaten Jember, 2020). The increase in population will undoubtedly increase domestic water demand. In addition to water demand for various sites, the demand for outflow components in this model is surface runoff and evapotranspiration. Surface runoff and evapotranspiration are the most significant components of the outflow. The value depends on the percentage of vegetation land cover-the more the vegetation cover, the smaller the runoff. The simple things that can be done to increase water demand for domestic, urban, industrial, and livestock are savings or efficiency measures. The application of saving demand for irrigation water can be made by applying the SRI method. The SRI method can reduce irrigation water demand by 25% with the SRI method (Subari et al., 2012).
However, to apply this method, a more in-depth study of the water demand for irrigation. The application of savings for irrigation water demand can also be made by changing cropping patterns. This is because the demand for irrigation water is the most significant demand than water demand in other sectors. This is supported by Kumalajati et al.(2017) research which states that agriculture is the largest user of water. Changes in paddy cropping three times a year can be replaced with palawija during the dry season. This can help reduce the water demand for irrigation, given that the water demand for palawija is much smaller than for rice. For groundwater preservation during the rainy season, a simple thing that can be done is to make infiltration wells in the house's yard. The same result was conveyed by Kahirun & Hasani (2017) regarding technical efforts in forest and land rehabilitation by applying soil and water conservation principles. This can help reduce surface runoff and prevent water from being discharged into water bodies.

Conclusion
The water balance of Mayang Watershed was observed for 18 years (2002 to 2019) using the WEAP model. The average annual precipitation that supplies the watershed is 1733.56 mm. The demand for water for each site tends to increase every year. The modeling results show that from 2002 to 2019, the water resources condition in the watershed can meet the demand from 3rd December to 1st May. The coverage value reaches 100%. However, from 2 nd May to 2 nd December, the demand for water from all demand sites was insufficient.
The highest first unmet demand occurred in October, with 0.67 million m 3 . The management of water resources, especially in terms of distribution during the rainy season and dry season, must be considered. It is also noted that the WEAP can simulate the main component of water balance in the watershed.

Conflict of Interest
The authors declare no conflict of interest.