1 Introduction

Water is necessary for the survival of humans, plants, and animals. It is one of the most valuable substances on the planet, required for drinking and other functions. It is also necessary for the healthy growth of crops. Access to safe and cheap drinking water is internationally acknowledged as a fundamental human right (United Nations 2019). Using this resource to live a quality and healthy life is one of a citizen’s fundamental rights. Water scarcity, poor water quality, irregular supply, and inadequate sanitation and hygiene practices have a significant impact on people’s well-being. According to Bogale (2020), by 2025, half of the world’s population would be living in water-stressed areas. Pakistan is likewise facing the threat of water scarcity (Ahmed et al. 2022). Pakistan’s per capita water availability was 2490 m3 in 1955, 1672 m3 in 1990 (Kahlown and Majeed 2003), and 1040 m3 in 2010. By 2035, it is expected to be reduced to 500 m3 (Zeshan and Shakeel, 2021). Per capita water supply is declining due to increased urbanization, inefficient use of available resources, and a lack of planning and information of soil data (type) Therefore, water conservation is critical to meeting future demand. Harvesting rainwater is an important technique for preserving the water that falls as precipitation. This technique has been used for over 4000 years while taking into account the spatial and temporal patterns of rainfall in the studied area (Khandelwal et al. 2019). Sustainable Development Goal 6 also emphasizes sustainable water management for all, whereas goal 6 (a) focuses on water harvesting (UN, 2018). There are three basic types of rainwater harvesting: in-situ rainwater harvesting, rooftop rainwater harvesting, and rainwater gathering by check dams, all of which can be used depending on specific needs and priorities. Before installing any rainwater harvesting structure, it is critical to assess the study area’s land appropriateness so that available resources can be used most effectively. Geographical information systems (GIS) and remote sensing are new technologies that can be used to find ideal sites based on a variety of criteria such as topography, land use, soil, slope, drainage, surface runoff, and so on (Balkhaira and Rahman 2021). Several studies have been conducted around the world to identify ideal areas for rainwater gathering using RS and GIS technologies. Scientists have indicated that these strategies are highly effective in the site selection process for good resource planning. The integration of Multi-Criteria Decision Analysis (MCDA) with GIS in West Bengal, India, to locate potential water harvesting and artificial recharge sites demonstrated the effectiveness of GIS technology (Singh et al. 2017).

The combination of MCDA, GIS technology, and hydrological models overcomes the significant challenge of finding possible rainwater harvesting (RWH) locations in Mashed Plain Basin (MPB), Iran (Toosia et al. 2019). Sayl et al. (2020) also employed advanced GIS technologies to conduct multi-criteria analysis while determining viable RWH sites in Iraq. Another GIS-based study conducted in Rajasthan’s Alwar district found seven RWH areas based on stated criteria (Tiwari et al. 2018). Similarly, the Soil Conservation Service Curve Number (SCS CN) is a popular method for estimating surface runoff required for locating RWH suitability locations (Mugo and Odera 2019; Sepehri et al. 2018). Assigning weights to various layers utilizing weighted overlay operation in GIS offers powerful outcomes on sites (Bera and Ahmad 2016). Similarly, recent study on RWH was carried out in Tajiskstan (Domullodzhanov and Rahmatilloev 2023) in the Kysylsu River Basin at variable altitude and rainfall depth. They proposed using local equipment to make storage tanks (100 m3) for household rainwater harvesting systems in various agricultural landscapes, harvesting and storing enough rainwater to satisfy the water shortage from May to October. Nepal is also facing severe climate change threats in Himalayas, therefore, a recent study (Tikhatri and Bhattarai 2023) proposed. RWH based on Rational method (rainfall intensity and runoff coefficient). They proposed RWH sites for the middle and southern plains, where precipitation ranged between 1,500 and 2,000 mm, as well as some high-altitude places in the north where precipitation was less than 1,000 mm.

Pakistan, a developing country, is not only adopting the installation of RWH structures, but is also promoting this practice due to its usefulness in water conservation. Ghani et al. (2013) proposed RWH sites near Potohar’s peak discharge locations using HEC-Geo-HMS and HEC-HMS. Siddiqui et al. (2018) conducted a study in Model Town, Lahore, to conserve rainwater from road runoff using GIS technology. Another study used RWH to artificially recharge the Lahore aquifer and found that the groundwater level in the city may rise to 3.54ft after each monsoon season (Hussain et al. 2019). Zia et al. (2021) employed the Analytical Hierarchy Process (AHP) in conjunction with GIS to find suitable rainwater harvesting locations in Lahore, and they found these techniques to be extremely useful for resource allocation and management. Siyal et al. (2018) used GIS to assess the position and size of water harvesting installations in Pakistan’s Karoonjhar mountainous terrain using CN numbers. Farooq et al. (2020) used ground information and GIS to estimate surface runoff and identified feasible rainwater collection sites in the hill torrent (rod-Kohi) region of D.I. Khan, Pakistan. Similarly, a recent study (Waseem et al. 2023) proposed RWH as sustainable solution by installing water filtration plants to cope with drinking water scarcity in Punjab (Lahore). They also employed an order stream network to identify probable ponding spots and the most appropriate places for underground water storage tanks.

Although the literature shows the use of RS and GIS in various studies conducted in recent years for the identification of RWH sites in plains, plateaus, and hilly areas, to our knowledge, no study has been conducted for the identification of potential RWH sites in a high-altitude area at the confluence of three mighty mountain ranges: the Karakoram, the Hindu Kush, and the Himalayan. As a result, this study aims to offer a baseline for identifying probable feasible sites for optimal rainwater harvesting in the Gilgit River Basin (GRB) using geospatial technologies. Because Gilgit-Baltistan (GB) sits at the crossroads of the China-Pakistan Economic Corridor (CPEC), there is an urgent need to conduct research and development in all areas. This location was also chosen because of its importance in terms of tourism, climate change, topography, and growing population pressure. Although snowmelt runoff from glaciers is the primary supply of fresh water in the study area, several water-related challenges have arisen in recent years. The majority of Gilgit people drink from the river, although the river is becoming contaminated as a result of human activity. The availability of clean drinking water is challenged not just by temperature and rainfall variations, but also by risky water management practices (Civil Society Coalition for Climate Change 2018). As a result, there is an urgent need to effectively maintain and utilize Gilgit’s available water resources through adequate planning, management, and allocation. This research will serve as a beginning point for researchers, development authorities, policymakers, and other stakeholders, including the GRB community, to solve their current water-related concerns.

2 Materials and methods

2.1 Study area

Gilgit-Baltistan (formerly Northern Areas) is located at the junction of three major mountain ranges: the Karakoram, the Himalayas and the Hindukush. The geographical location of this region is very significant, and it is situated at the extreme North of Pakistan. It covers an area of 72,496 km2 and shares its border with China to the Northeast, Afghanistan to the North, Khyber Pakhtunkhwa to the West and Indian held Jammu and Kashmir to the Southeast. Administratively, GB is divided into two divisions and ten districts. Gilgit City is the capital of GB. According to 1998 census, the province had a population of 884,000, whereas, it had risen to 1,492,924 in 2017 (Planning and Development Department-Government of Baltistan 2020) at a rate of about 3.6% per annum. Climatically, July is the hottest month in GB with mean monthly temperature range (> 14oC to 40oC). January is the coldest month with a mean monthly temperature range (-7.6oC to 11.1oC). Therefore, the mean monthly maximum temperature is 26.7oC and mean monthly minimum temperature is 3.4oC. GB receives mean annual precipitation of 208 mm (Javed et al. 2020).

The study area (Fig. 1) is showing GRB that is located between 72.34 and 75.49 oE and 35.41 to 37.29 oN and covers an area of 27,273 km2 of GB. The elevation ranges from 1265 to 7809 m above mean sea level. Two major rivers Gilgit and Hunza along with their many tributaries (nullahs/streams) flow in the GRB. These water channels carry huge amounts of water and fulfill the needs of residents. GRB covers an area of 27,273 km2 of GB computed at Alam Bridge stream flow station. The elevation ranges from 1265 to 7809 m above mean sea level (Fig. 1). Glaciers cover around 7% of the region, with precipitation varying from 50 mm to 350 mm and snow cover ranging from 3% to 90% during summer and spring seasons, respectively (Hasson et al. 2014). Its urban area is Gilgit city which is the center of the Gilgit District. It had population 79,349 including Danyore and Sakwar according to 1998 census whereas, its projected population for year 2025 is 331,000 (Annandale, 2014). Geographically, Gilgit City is located between 74.32 to 74.53oE and 35.83 to 35.98oN in the Karakoram Mountains and occupied an area of around 20,000 Acres or 85 sq-km with elevation 1500 approximately (MM Pakistan (Pvt) Ltd., 2020). It is among the major hubs for all mountaineering expeditions in the Northern Areas of Pakistan. Trading, mining and agriculture are main occupations in this city. Two major rivers Gilgit and Hunza along with many tributaries (nallahs/streams) flow in the Gilgit City. These water channels carry huge amount of water and fulfill the needs of residents. The major nallahs in the Gilgit City are Jutial, Minawar, Sakwar, Kargah and Danyore. Besides, there are five drinking water supply complexes available for the storage of water in the Gilgit City (Karar, 2011).

Fig. 1
figure 1

Study area map of Gilgit River Basin showing the elevation, water channels, stream gauge and weather stations

Full size image

2.2 Data collection

The SRTM Digital Elevation Model (DEM) was obtained from United States Geological Survey (USGS) for watershed delineation whereas, coordinates of delineation point at Alam Bridge-Gilgit District were acquired from WAPDA. The land use/land cover (LULC) data of Gilgit watershed was acquired from ESRI. Hydrological soil data of Food and Agriculture Organization (FAO) was used to classify diverse soil groups. Finally, rainfall data of last 22 years (2001–2022) of Gilgit watershed was obtained from Pakistan Metrological Department (PMD) and Water and Power Development Authority (WAPDA) for computation of annual average precipitation (Table 1).

Table 1 Details of data sources and their spatial resolution
Full size table

2.3 Data preparation and processing

Watershed was delineated in the licensed version of ArcGIS 10.8.1. The Arc Hydro extension was used to carry out all the typical steps. Primarily, fill sink operation was applied on SRTM DEM to avoid depressions in the raster. The procedures of Flow Direction, Flow Accumulation, Stream Definition, Stream Segmentation, Catchment Grid Delineation and Catchment Polygon Processing were executed to isolate the sub catchments in the study area. Drainage lines and drainage points were also computed and finally, a point was defined to demarcate the watershed boundary and calculate its area. To prepare a LULC map of the area of interest, the downloaded global LULC was clipped in ArcMap using Gilgit watershed boundary. Similarly, the hydrological soil map was organized for its further utilization. Table 2 explains the variety of hydrological soil texture with its characteristics as described by United States Department of Agriculture (USDA) 1974 (Sekar et al. 2017). Lastly, the rainfall data of 11 climatic stations including Gilgit, Bunji, Gupis, Krimabad, Khotpass, Khunjerab, Naltar, Shendure, Ushkore, Yasin and Ziarat from 2001 to 2022 was processed to determine annual average rainfall of the respective stations. The location, elevation and mean annual precipitation are provided in Table 3. Joining spatial data of climatic stations with rainfall value led to calculate average rainfall of the entire basin. There are three methods available for measuring annual rainfall depth of the basin based on their certain characteristics, namely: Arithmetic average method, Thiessen polygon method and Isohyetal method. The most imperative method among all is isohyetal method (Al-Timimi et al. 2020). This method was employed to compute the annual average rainfall of the basin. For this purpose, a model was built using ArcMap in which point data was interpolated to generate a surface using rainfall value. Filled Contour was created with 50 m interval to calculate the area in km2 between two consecutive isohyets. Similarly, average rainfall was computed by taking an average of two successive isohyets and it was multiplied with the corresponding area between them to determine the rainfall volume. Equation (1) was followed to calculate the average annual rainfall of the basin for the last 22 years Fig. 2. Precipitation data for all stations was available from 2001 to 2022. 

$${P}_{ave}=\left({\Sigma }\text{A}1\text{P}1\right)/\varSigma A1$$
(1)
$${P}_{ave}=Average\,Precipitation$$
$${\Sigma }A=Total\,area\,of\,the\,Basin$$
$$P1=Respective\,Gauge\,Station$$
$$A1=Area\,between\,two\,successive\,isohyets$$
Table 2 Hydrological soil group classification based on global hydrologic soil groups (HYSOGs250m) (Ross et al., 2018).
Full size table
Table 3 Location, elevation and mean annual precipitation at rainfall stations in Gilgit River basin
Full size table
Fig. 2
figure 2

Time series graph of isohyetal based spatially distributed mean annual precipitation of Gilgit River basin

Full size image

Figure 3 explains the layout of methodology in detail. The thematic layers were prepared individually to run the complete process. Four parameters including drainage density, surface runoff, slope and LULC were selected to identify rainwater harvesting sites in GRB based on the criteria used by (Mahmood et al. 2020). Many studies conducted for the identification of RWH sites have emphasized the importance of these parameters. Some studies have selected the same four layers for the identification of RWH sites, e.g., in the northern part of Iraq (Buraihia and Shariff 2015), Wadi Numan watershed of Saudi Arabia (Al-Ghobari and Dewidar 2021) and District Kohat, Pakistan (Farooq et al. 2022). However, some studies have used these in combination with others parameters, e.g., the research carried out in Kakreza watershed, Iran for determining potential RWH sites (Karimi and Zeinivand 2019), detecting artificial recharge zone in Anantapur district, India (Rajasekhar et al. 2020), assessing potential water storage sites in the western part of Iraq (Hashim and Sayl 2020), and MCDA for effective RWH in the northeast of Iran (Toosia et al. 2019).

Fig. 3
figure 3

Methodological flowchart for the selection of potential rainwater harvesting sites

Full size image

The drainage density, total length of a stream per unit catchment area, was calculated using Eq. (2). The drainage network developed during the watershed delineation was spatially joined with sub catchment layers having their calculated area in km2. Length of total streams within a respective sub catchment was determined in kilometers and the drainage density of each sub catchment was also calculated accordingly. Mathematically, drainage density is computed using the equation below (Cevi et al. 2019) in which L is representing the length and A is the total area of a sub-catchment.

$$\text{Drainage Density}\,=\,{{\sum }_{i=L}^{n}L/A}$$
(2)

Slope is the steepness at each cell of a raster surface. The higher value of slope indicates steepness and lower value indicates flatness of the terrain. Therefore, processed SRTM DEM free from sinks was used to determine slope of the study area using hydrological tools. Surface runoff computation was a challenging task which was achieved using the Soil Conservation Service (SCS) Curve Number method, originally developed by USDA. This method is widely used for the estimation of surface runoff, and it is explained in National Engineering Handbook (NEH, 2017) section in detail (Mockus 1964). The Gilgit River Basin is a snow-fed basin with a small glacier melt contribution (around 9%) (Latif et al. 2020) to the total runoff, therefore the runoff under snowmelt can be computed from the rainfall depths and direct runoff using CN/SCS methods (Hejduk et al. 2015). There are multiple parameters for determining surface runoff, e.g., landuse/land cover, soil type, topographic details and moisture condition of the area and they all lead to the generation of CN grid. This method is based on water balance equation and surface runoff can be calculated using Eq. (3)

$${Q}={(P-Ia)^{\exp 2}}/(P-Ia+S)$$
(3)

where Q is direct runoff (mm), p is total rainfall of the area (mm), \({I}_{a}\)is initial abstraction (mm), and S is potential maximum retention (mm). It is true only when P is greater than or equal to\({I}_{a}\). It is linked with two assumptions: one postulation is that the ratio of actual surface runoff to actual rainfall is equal to the ratio of actual infiltration to maximum potential retention. Second postulation directs to initial abstraction \({I}_{a}\) (Topcuoglu et al. 2021) which is

$${I}_{a}=0.2\,S$$
(4)

thus;

$${Q}={(P-0.2\,S)^{\exp 2}}/(P+0.8S)$$
(5)

Whereas, S is calculated using CN computed with combination of parameters

$$S=\raisebox{1ex}{$25,400$}\!\left/ \!\raisebox{-1ex}{$CN$}\right.-254$$
(6)

CN is dimensionless and has a range from 0 to 100. The maximum value 100 shows the maximum runoff generated from rainfall, whereas the minimum value of CN shows that all rainfall is absorbed by the surface (Krois and Schulte 2014). To calculate CN value per pixel in the area of interest, HEC-GeoHMS (Geospatial Hydrologic Modeling Extension) has a built-in function that takes inputs of HydroDEM, merged layer of soil & landuse and CNLookUp table.

In the next step, Landuse, drainage density, runoff and slope datasets were classified for performing weighted overlay analysis to find suitable sites for RWH. The relative weight of surface runoff, slope, drainge density and landuse were 48%, 30%, 15% and 7%, respectively, which were based on the criteria adopted by (Mahmood et al. 2020; Buraihia and Shariff 2015) in Table 4. Finally, the individual artificial structures e.g., check dams, farm ponds, or percolation tanks were proposed on the already found medium to highly suitable areas for rainwater harvesting. Three layers including landuse/land cover, soil type and slope were reclassified for each individual artificial structure and the computations were carried out using ‘raster calculator’ in ArcGIS software (Table 5). (Al-Ghobari and Dewidar 2021; Mahmood et al. 2020; Khalid et al. 2017; Adham et al. 2016) also used these layers for proposing rainwater harvesting structures in their study areas. In Pakistan’s arid regions or plain areas, the effects of soil physicochemical parameters under various land use types were investigated by a recent study (Yaseen et al. 2023). They classified the forest, cultivated, and grazing regions using soil sampling at depths ranging from 0 to 60 cm, with a 20-cm gap for the selected strata. The soild structure is an important parameter for slecteing RWH sites based on the water holding capacity. Forest layers contain more sandy soils (67% sand particles) compared to cultivated land (37% clay particles).

Table 4 Selected Parameters and their scale value (Mahmood et al. 2020)
Full size table
Table 5 Criteria for site selection for RWH installtion structure
Full size table

3 Results and discussion

Figure 4 depicts a precipitation map based on its mean annual rainfall which showed that the western parts of the study area received more rainfall than the eastern parts. Nalter site has recorded the highest rainfall whereas the lowest precipitation in study area were noted at Gilgit weather station. Figure 5 illustrates the zonal pattern of (a) Landuse and Landcover, (b) Hydrological Soil groups, (c) Curve Number, (d) Surface Runoff, (e) Slope and (f) Drainage density in Gilgit Watershed. Consequently, the parameters used (Slope, Landuse/land cover, drainage density and surface runoff) were imported into the model that estimated the possible sites for the installation of the rainwater harvesting system. The land use types of the study area was divided into six major categories: water and snow, built up areas, forest, agriculture, and bare land. Almost half (nearly 44%) of the study area was identified as non-productive land which can be developed as rainwater harvesting storage structures (Mugo and Odera 2019; Aghaloo and Chiu 2020). However, the percentage of area which was covered by building, water, and snow was about 27% which was obviously non-potential for RWH structures, and therefore had the smallest weighting factor in the suitability analysis. According to the GIS data, 29% of the agricultural land was used for crop production, basic grassland field, and shrubs mostly in the northwest and southwest areas. In the Fig. 5b, three main kinds of hydrological soil in the area, namely B, C and D are shown. Each soil type was listed previously in Table 2 that follows the USDA classification (USDA, 1974). Class C (moderately high) was seen mostly in central and northeastern portions of the Gilgit basin with total area of 8496 km2. Conversely, group D belonging to the soil was located near the edge of the study area, in an area of 3332 km2. This category D, which was significantly lower in infiltration rate and higher in surface runoff generation, was well adapted for the rainwater harvesting structure constructions. The RWH data used for only 37% area mostly covered by water, snow and ice was not useful. The dimensionless Curve Number computed by combination of land use together with hydrological soil groups showed a range from 58 to 100. (Fig. 5c) Using isohyetal method and it was calculated for the average precipitation in 2022  for the basin and the value became 472 mm. Farooq et al. (2020) conducted a qualitative study to identify surface runoff potential that can be used effectively for management of torrent issues with water harvesting sites of selected catchments of D.I. Khan District, Pakistan. According to their data, the investigation area exhibited a mean annual precipitation of 292 mm, while; in case of our study (GRB), the area received a greater amount of mean annual precipitation due to the extent of snow as solid precipitation, as revealed in Fig. 4. Mathematical modeling of curve number values against the average precipitation values has generated runoff of about 306 mm to 472 mm as depicted in Fig. 5d. The range of slope percent values was from 0 to 77% in GRB.

Fig. 4
figure 4

Spatially distributed mean annual precipitation rainfall map of Gilgit River basin

Full size image
Fig. 5
figure 5

Spatial variation of multiple parameters (a) Landuse and Land Cover, (b) Hydrological Soil groups, (c) Curve Number, (d) Surface Runoff, (e) Slope and (f) Drainage Density in Gilgit River basin

Full size image

Among the many slope values, the highest value was assigned to the group with slopes less than 5% and the lowest value – which was totally unsuitable for rainwater harvesting structures – was given to groups having slopes which were above 27.7%. According to the findings, only 19% of the area having slopes less than 15% was capable to develop rainwater storage structures as suggested in Table 4 (Adham et al. 2016). Furthermore, the drainage density ranged from 0 to 0.72/km2 in GRB (Fig. 5f). The spatially higher drainage density values were given low ranks and the spatially lower drainage density values received higher ranks in the site performances for suitability maps.

After overlaying all these thematic layers in GIS environment using weighted overlay operation, the resultant output layer was categorized into four major classes: RWH is high, medium and low suitability by some places and by some places is unsuitable. This is what the study derived in the shape of a rainwater harvesting sites suitability map shown in Fig. 6. According to that information, only 1147 km2 (4%) would have been highly suitable while 40% of the whole research area would have fallen into the interval called medium suitability. Thus, 44% area of the Gilgit watershed can be used for rainwater harvesting (Fig. 7). This is the area where the runoff reached its maximum level, and the soil had low infiltration rate and the land was barren and agricultural. This appropriate location was located in the central part of the study area in lower Gilgit city and dispersed very small areas were detected in different parts of the GRB. According to the study 19% of the area that was considered can be placed water conservation measures that were suitable and it was predominantly found on the western side of the GRB.

The existence of highly snowy mountains mostly in the eastern part of GRB was totally against effective water conservation plans.

Fig. 6
figure 6

Suitability map for potential rainwater harvesting sites

Full size image
Fig. 7
figure 7

Distribution of land suitability(%) for potential rainwater harvesting sites

Full size image

The fuzzy gamma overlay using slope, stream order, precipitation, evaporation, Nomalised Differnce Vegetation Index (NDVI), soil and road datasets applied by Alwan et al. (2020) was stated as paramount which comprised 56% of Maysan Province, Iraq, suitable for rainwater harvesting purpose. Agricultural areas were the main targets for the construction of rainwater harvesting sites and towns’ agglomerations were avoided for this construction purposes. An elevation variation zone ranging from 10 to 264 m above mean sea level (amsl) and a slope zone of 0-54% characterized their study area. Conversely, this study entailed weighted overlay analysis that concentrated on the barren and agricultural land for potential rain harvesting sites. The mean of elevation of GRB is about 4500 m amsl and up to 75% of the slopes; hence only 4% of the catchment was suitable for the point sources of RWH as a site. The research carried out in Toughda catchment, Morocco with a mean altitude of 1500 m used a combination of GIS, Analytic Hierarchy Process, and remote sensing approaches and declared 5.93% of the total area mostly suitable for holding structures, such as the hill dams (Ouali et al. 2022). (Buraihia and Shariff 2015) conducted a study in an area less pronounced gradient in Kirkuk, Iraq (average altitude of 331 m). We used the same parameters and influence values as they did, however,  these authors found, 71.6% area being suitable for rainwater harvesting structures, such as farm ponds and check dams in agricultural land. Utilizing the same layers (slope, landuse/landcover, surface runoff, drainage density), Al-Ghobari and Dewidar, 2021 analyzed watershed of Wadi Numan semi-area of western region in Saudi Arabia (average elevation of 941 m), which has 40% of the study area good for conserving rainwater including farmlands and bare soils. Also, Srivalli and Manjeet (2019) determined weighted overlay for seven factors such as slope, landuse/landcover, soil, geomorphology, geology, water transmissivity in soil, and runoff on cultivable and barren land of Gadela watershed in the Udaipur District, Rajhistan, India by reporting 97% area with high to medium suitability for potential RWH sites. However, we confirmed the high to medium suitability of nearly half of this basin for water preservation due to the fact that huge snow cover exists. According to data obtained by Mahmood et al. (2020) the same characteristics (gradient, LULC, surface runoff and draiange density) and their influence to set possible water harvesting sites have been identified with the help of remote sensing and GIS techniques in Kohat District, Khaiber Pakhtoonkha, Pakistan. The result of their study was that just 17% area was having poor suitability for RWH while in our study almost 56% of area was having poor suitability for RWH. The main difference in their research and our result is due to the disparity in their average elevations of the Kohat district which is 489 m and the GRB which is 4500 m. Furthermore, another imporatant aspect of differenc is attributed to snow inclusion in the precipitation in GRB. The unsuitable area will be further minimised after distinguishing snow from precipitation.

3.1 Recommended structures at RWH sites

Furthermore, the landuse/land cover, soil type and slope layers were used to propose the individual artificial structures e.g. check dams, farm ponds, or percolation tanks on the medium to highly suitable (44%) areas for rainwater harvesting. The results showed that only 3.5% of the already found medium to highly suitable RWH areas satisfied the conditions of C type soil having moderate runoff & moderate water transmission, barren land and slope < 15% for the development of check dams. 1% of the medium to highly suitable RWH areas were appropriate for the development of percolation tanks because it had hydrological soil type as sandy clay loam or clay loam, barren land and slope < 10%. About 0.75% of the medium to highly suitable RWH areas were proposed for the development of farm ponds with soil type (C and D) having slow water transmission rate, agricultural land and slope < 5%. These proposed locations for check dams, farm ponds, or percolation tanks are shown in Fig. 8. Mahmood et al. (2020) also proposed artificial RWH structers in district Kohat, Pakistan and found 4.21% of the total study area suitable for percolation tanks, 3.21% for farm ponds and 3% suitable for the development of check dams. Whereas, Al-Ghobari and Dewidar (2021) proposed less than 1% of their study area suitable for the development of farm ponds due to semi-arid conditions of Wadi Nu’man watershed in the Western region of Saudi Arabia.

Fig. 8
figure 8

Recommended structures at potential rainwater harvesting sites

Full size image

Athough, a large amount of water coming from rivers, nullahs, glaciers in addition to rainfall is currently available but a number of water related issues already exist in the Gilgit River Basin. As Gilgit River Basin (Pakistan) is connected with China and both countries are working on CPEC, a large economic bustle with an increased incoming population will soon put pressure on the available water resources if they are not managed properly. The use of RS and GIS technologies in the optimal site selection process for rainwater harvesting showed that water resource management could be carried out in an effeicient and optimal way using geospatial technologies. This study can be helpful for the Gilgit Development Authority and other concerned departments in their planning phase by providing suitability sites baseline for study area.

4 Conclusion

The aim of the study was twofold: locating potential places where water collects from the rain and selecting only the suitable ones, which is the Gilgit River basin at the high altitudes. The study location is of great importance due to the world-famous glaciers, Tousrim, and China Pakistan Economic Corridor project. In GIS environment, points of potential rainwater cavitation were determined considering slope angle, drainage’s density, land use and surface runoff. The amount of runoff was determined by using the SCS-CN method which taking into account use of land, the types of soils, precipitation, and the DEM. The analysis showed that about 44% of the infertile lands could be used to develop RWH structures, although only adding 4% could be considered very suitable for such development. 37% of the study area’s area is absolutely impossible to water harvest structures just because it is almost completely covered with snow and water. The existence of storm drain points or discharges from water bodies showed that the building of the RWH systems were possible as there were enough locations available from where the water could be collected to be used when required. The Gilgit River catchment receives sufficient waters from rainfalls, nullahs, rivers and glaciers at present. Although CPEC falls under the collaboration between China and Pakistan, Gilgit Baltistan shares a border with China, which induces the rise of some commercial activities in the CPEC’s route. It obviously will bring in people not only from within the country but also from other parts of Pakistan. Therefore, if the construction of rainfall harvesting or newly developed reservoirs does not take place, the existing water sources will become more pressurized and not be enough for the current consumption. Using geospatial techniques in study, convenient RWH sites in the Gilgit river basin, were identified, which facilitates the implementation of plans by Gilgit Development Authority and other departments in Gilgit region, Pakistan, in areas of planning, resource allocation, and responsible management of available water resources.