The pre-selection of suitable sites for small underground dams in arid areas using GIS (A case study in Yazd_Ardakan watershed) Jalal Barkhordari 1. Faculty member of Agricultural Natural Resource Research center P.o.Box 89165-517 Yazd, Iran E-mail: Barkhordari@alumni.itc.nl Abstract: A good location for underground dams meets a set of local and environmental criteria. Not all of these criteria are known and are nowadays partly underestimated during sitting of locations, resulting in minor efficiency of constructed dams. In most cases only large-scale database maps as geology, topography is available, which does not provide adequate information on detail environmental parameters. For this purpose, remote sensing techniques can supply large amount of data with high spatial and temporal resolution, and are therefore a very useful mapping tool especially in areas where very little information is available such as most developing countries. In this work, we present a methodology for the assessment of the suitability of sites for the installation of small underground dams, a technology widely used for water harvesting in arid climates. The selection criteria are defined both in a qualitative and quantitative way, and are based on a territorial analysis using satellite data and hydrological and climatological information that are easily and freely available. The methodology, applied to Yazd_Ardakan watershed in central Iran, allowed the individuation of 35 sites for subsurface dam and 114 sits for sand storage dam passed the proposed selection criteria. The validation results showed, Most of underground dam sites (46.7%) were found within the moderately suitable followed by highly suitable with 26.6%. The fact that most of predicted underground dam sits were found within marginally to highly suitable this indicates that the developed GIS and decision tools can reliably be used to predict potential sites for underground dam sites. Keywords: Site selection, Small underground dams; Morphological parameters remotely sensed, GIS, RS 1. Introduction Water resource availability in arid lands is mostly connected with the characteristics of shortage and periodicity of precipitations which determine runoff during only short periods of year. These problems, with large evaporation from crops which generally characterize the arid zones, impede the formation of perennial streams capable of satisfying environment and population s requirements. At the same time, the soil loses its infiltration and storage capacity as a result of a long period of drought and wind erosion (Forzieri et.al [2008]). Water use in arid lands, often located in developing countries, is limited by lack of financial resources and suitable technology in addition to physical conditions. These considerations support the use of surface and small underground dams for the utilization of water resources in developing countries with arid conditions. These kinds of structures present favorable characteristics both in terms of efficiency and simplicity of realization (Hansson and Nilsson [1986]; Cavalcanti et al. [1999]; Cavalcanti and Resende [2001],Forzieri et.al [2008]. The effective realization of a dam requires detailed surveys of site to evaluate the underground characteristics, fundamental for the sizing and functionality of the structure. Such surveys require large efforts that rarely allow a regional scale application hence hampering even feasibility studies in the case of large dams (Tabatabaii yazdi [2002]). We need to consider that the planning for water resources management and exploitation must be approached at regional scale, but often with limited financial resources. The acquisition of geographical and environmental information is then bound to a low cost framework. Before the start of the project engineering, however, the consideration of how an intervention, of any type, is entered in the surrounding context cannot be neglected (Nilsson [1985]). In a study, the groundwater artificial recharge suitable area of Silakhor, Borujerd of Iran rangelands was investigated and extracted using Geographic Information System (GIS) and Analytic Hierarchy Process (AHP) method. The layer information of rainfall, soil, lithology, slope, land use and fault components were prepared and weighed by applying AHP method and were used for artificial recharge site selection. Aquifer artificial recharge maps showed that in Silakhor catchment 14%, 64.8% and 21.2% of the area of the catchment had high, moderate and low capability for artificial recharge, respectively. The Vol. 6, Issue 1, March 2015 18
most suitable areas were located in the low land areas (Mehrabi et.al [2012]). A suitable low-cost approach may be based on the organization of the survey in two main phases: the first at large-scale and limited resolution where a number of possible sites of interest are selected, the topic of this study, and the second where these sites are object of a point survey with larger accuracy (Barkhordari [2013]). In this study we present some site selection criteria for the realization of water reservoirs, which use information acquired prevalently by satellite platform. Through these criteria it was possible to remotely evaluate the suitability of sites for the construction of small underground dams. The methodology has been applied for the pre-selection of sites for the installation of small underground dams, in Yazd_Ardakan watershed. 2. Underground dams An underground dam is a structure that obstructs the natural flow of ground water and stores water below the ground surface. There are basically two types of ground water dames; i) Subsurface dames ii) Sand storage dames (see Figs. 1 and 2). A subsurface dam is constructed below ground level and arrests the flow in a natural aquifer. Whereas sand storage dam impounds water in sediments caused to accumulate by the itself (Hanson and Nilsson [1986]). Figure 1: Principles of a subsurface dam (Irura [2006]) Figure 2: Principles of a sand storage dam (Barkhordari [2013]) Vol. 6, Issue 1, March 2015 19
3. Materials and Methods 3.1. Study area Yazd-Ardakan watershed has area about 1595000 ha which is located in Northern part of Yazd Province and included 12.3% of 13 million ha area of the province. This area has been extended between the longitudes of 52 o 57' to 54 o 59' and latitudes 31 o 13' to 32 o 48' of Iran central plateau (Figure 1.1). This area is surrounded by mountains and the general slope is from North-West to South-East to Siyah-kouh plateau. The region climate is dry and cold, in way of modified Demarton method. Precipitation average is 62.1 mm and ETP is 3483mm. Yazd-Ardakan District have some main sub-basin such as Pishkouh and Khezrabad catchments that are more important in securing water (Barkhordari [2013]). 3.2. Methodology Figure 3: Location of Yazd _ Ardakan watershed Selecting suitable sites for underground dam structures is a multi-objectives and multi-variables. Here, objectives are suitable sites for subsurface dam, sand-storage dam and non suitable area. The Variables are main Criteria's which are presented in scientific references for suitability underground dam sits as landuse, soil, stream slope, runoff potential,, geology, and stream order maps). The variables are overlaid with some diction roles for selecting the objectives in GIS (Barkhordari [2013], Nissen-Petersen [2000]; Tabatabaii yazdi [2002]; Forzieri et.al [2008] & Irura [2006]). The parameters used to evaluate the suitability of sites for the installation of underground dams are described below. Most of the reference parameters are evaluated by remote analysis and are based on the interpretation of satellite images and large-scale available cartography. Other parameters are evaluated by in situ surveys, in the region of study area, which allowed improvement of the proposed selection criteria. Combined with information obtained from the field this can give valuable information about the important factors controlling the efficiency of an underground dam. The main goal of an underground dam is to increase the volume of water and (coarse) sand in a riverbed in dry land areas where episodic shortages of water are common. Several reports describe Vol. 6, Issue 1, March 2015 20
the basic local and environmental criteria concerning the successful construction of an Underground dam In this work, The selection criteria are defined both in a qualitative and quantitative way, and are based on a territorial analysis using satellite data (images, digital elevation models) and hydrological and climatologically information that are easily and freely available. The suitable site for Underground Dam was prepared by integrating different thematic maps in the GIS environment. The variable maps were integrated using cross method by Decision rules (DSS).Then for Validation of the selected Underground Dam sits were assessed using information obtained from the field survey(10 point as Sand Storage Dams and 5 point as Subsurface Dams) and the generated suitability map. 4. Results and discussion 4.1. Geomorphic characteristics Stream characteristics: Stream network of study area are created using the topographic maps (Scale 1/50000) and Digital Elevation Model in GIS program. Then use it for calculation Stream Network Ordering, Stream Slope and the length of the reach of wadi contained in the influence area of the barrage in study area (see Fig. 4). Figure 4: Miankouh River, showing digitized black polygon delineating the influence area of barrage and red line delineating the stream length Narrows width. The localization of narrows near village is obtained by comparative analysis of available informative layers: land cover map1:250,000 (Barkhordari &Vardanian,2011), satellite images and digital elevation model (DEM).The estimate of narrows width (L) is obtained by visual interpretation procedure elaborated by satellite images in GIS environment(see Fig. 4). Vol. 6, Issue 1, March 2015 21
Influence area of barrage. The influence area corresponds to the upstream narrows surface which, after the realization of the barrage, will benefit from the increment of groundwater resource (FAO [2001]). The estimation of this surface is obtained by visual interpretation of satellite images in the visible bands (LANDSAT 7 ETM+) in GIS environment (Fig. 4). Available Sandy Rivers: In this part is created sand availability map of study area using GIS and satellite images (Landsat 7 ETM+ image for Jun 2009) and also, during filed work have been collected 60 sediment surface samples in dry rivers on Jun to August. Median grain size of the sediment (D 50 in mm) was determined in laboratory. The 40 samples have been used for calculation of relation equation between satellite reflectance and samples median grain size and 20 samples for validation of relation equation. Different soil types have different grain sizes and spectral responses. The size and arrangement of the soil particles have an impact on the soil reflectance. Obtaining a relationship between the grain size and the spectral response of the sediment samples taken in the study area validates whether pixel values are correctly identified as sand. Samples used for the validation are from the Pishkouh and the Khezrabad sub basins. The width of the riverbeds at these locations approaches the spatial resolution of the ETM + images, allowing a more detailed study on the grain size for sands. The spectral response to grain size shows the largest variation in the Bands 3 and 4, varying from Digital Number (DN) 50 to DN 147 and from smaller median grain size (107 μm) to larger median grain size (673 μm). Band 2 shows the smallest variation. The best relationship between DN and D 50 is obtained for band 4, with an R 2 of 0.66(fig.5) (Barkhordari [2013]). Figure 5: Multiple regression analysis of field samples Faults: Faults map of study area have been prepared by Geology map and Satellite images of study area. Attentions to the faulty area are not suitable for underground dam construction then created 500 meter boundary fault lines. Bedrock depth and substrate porosity: After investigation of the alluvium properties and rock lithology for permeability of the reservoir bed rock have been classified in three classes: high, medium and low. Permeability classes lithology units are presented Table (1) below. Based on engineering properties, the alluvial of river deposits are high permeability and porosity of over 30% that are appropriate for underground dam reservoir. 4.2. Vegetation cover NDVI: The normalized difference vegetation index (NDVI, February 1988) is an index of vegetation cover and it is used to quantify vegetable biomass and vegetation health. The index was determinate by multispectral analysis of satellite imagery in the red and near infrared bands. 4.3. Pre-existent infrastructures and villages The other information used for the pre-selection of suitable sites for small dikes are given by the presence of road infrastructures, obtained and localization of villages and other human settlements by available land cover map Vol. 6, Issue 1, March 2015 22
1:250,000 (Barkhordari& Vardanian [2011]).The other useful parameters, not evaluated in this study due to the lack of data, are the potential user, i.e. the number of people which could benefit from the barrage, and by aquifer dislocation near the narrows. 4.4. Climatologically characteristics A rainfall-runoff water balance model at a meso level has been used to estimate spatial runoff variation and the total catchment runoff produced in a watershed in a monthly time step. Such a model investigates the contribution of surface runoff and groundwater flow to the total catchment runoff. The runoff potential map has been classier into low, moderate, high and very high runoff potential zones (Fig.6). Most of the areas 88.24 % are dominated by low and moderate runoff potential zones that can be suitable zone for selecting subsurface dams. Whereas high and very high runoff potential zones cover an area of 11.76% of the study area that can be suitable for sand storage dams (Barkhordari& Semsar Yazdi [2015]). Figure 6: Annual Potential runoff classification of study area Geology: The geology features are important in quality and quantity of water storage areas. Water quality is an important as the quantity criterion for water resource projects. If there are sources of pollution, such as salt domes and sediments or large industrial plants, etc. These areas are located in the first stage are eliminated. In this study, determined distributed underground water quality(fig.7) by using Yazd Regional Water Office data point to that were measured the electrical conductivity of water wells, qanats and springs scattered in several locations that were good, were preparation in study area. Vol. 6, Issue 1, March 2015 23
Figure7: Distributed underground water quality map of Study area After investigation of the alluvium properties and rock lithology for permeability of the reservoir bed rock have been classified in three classes: high, medium and low. Permeability classes lithology units are presented Table (1) below. Based on engineering properties, the alluvial of river deposits are high permeability and porosity of over 30% that are appropriate for underground dam reservoir. Table 1: Classification permeability of Geology formation Permeability High moderate Low Geology formations Q3- Qal- Qc-Qf1-Qf2-Qt2-Qt3-Qtr gb-kr-m cl- M am-m cs- M m-m rm-q pl(c)- Qt1-Rc C b- Cs-C ZI- Cb t- CO m- gr- KU(fI)- M I- M m3- M mg- M ms- M t- P d-p r- PC k Faults map of study area have been prepared by Geology map and Satellite images of study area. Attentions to the faulty area are not suitable for underground dam construction then created 500 meter boundary fault lines. Usually underground dam are construct up or near of village, farm lands, agriculture wells, springs or qanats. Then created location of used water resources of study area, too. 4.5. Selecting Suitable Sites for underground dam Structures by DSS The methodology proposed is based on a multi-attribute decision method. This process is constituted by a first phase of sites identification, by a second phase of qualitative selection based on geomorphic and functional aspects and by a third classification phase based on the estimation of barrage (underground dams construction)effectiveness. Technical guidelines have been suggested by Nissen-Petersen [2000]; Tabatabaii Yazdi [2002]; Irura [2006] and etc Vol. 6, Issue 1, March 2015 24
for selecting suitable sites for underground dam construction on many variables like as slope and ordering streams, land use, potential of sand availability in riverbeds, geology, stream, runoff potential, proximity to utility points, etc. These guidelines are used as an expert knowledge in the developed expert classifier shell and discussed in the following sections. The suitable site for Underground Dam was prepared by integrating different thematic maps in the GIS environment. The variable maps were integrated using cross method (Table 2). Table 2: Decision rules for site suitability selection Underground dam types Subsurface dams Sand dams storage Stream order Stream Slope (%) Runoff potential 4-5 <5 Medium to low 2-3 >5 High to medium Geology No faulty & low permeability No faulty & Low medium permeability to Potential of Sand deposit High to medium Medium to low Location (land use) Near of village,farm lands or qanat Near of village,farm lands or qanat 4.6. Sites identification The first important characteristic to localize a potentially suitable site for an underground dam is constituted by a reduction of the cross section of wadi (a narrows). During the construction phase, these characteristics allow to use the natural blockage of rock layers, both at the sides and in depth, as support at the barrage structure. This allows the reduction of barrage dimensions and the realization costs (Nilsson [1985]; Nilsson [1988]). The identification phase is conducted by visual interpretation of satellite images and analysis of large-scale cartography (Fig. 8). Figure 8: Suitable sites for underground dam structures in study area It was found that suitable sites for underground dam in the Yazd_Ardakan district covers an area of nearly 0.5% of the study area whereas the rest of the area was assumed as unsuitable site for underground dams.the areas which are suitable for underground dam structures such as subsurface dams and sand storage dams were suggested on Vol. 6, Issue 1, March 2015 25
figure 8 as underground dam Suitability map for the study area. On basis of suitable and unsuitable sites for underground dam structure map has been prepared. The suitable/unsuitable map which was prepared from integrating thematic layers as per the above laying decision rules. The suitability model generated a suitability map for Subsurface Dams and Sand Storage Dams. The methodology, applied to Yazd_Ardakan watershed in central Iran, allowed the individuation of 35 sites for subsurface dam and 114 sits for sand storage dam passed the proposed selection criteria. 4.7. Method Validation Validation of the selected Underground Dam sits were assessed using information obtained from the field survey(10 point as Sand Storage Dams and 5 point as Subsurface Dams) and the generated suitability map. For the purpose of the validation, some selected sits as random samples were assessed and classified them based on suitability. The results are shown in Table4. Table 4: Comparison of underground dam sits actual locations and predicted ones Underground dam Highly Moderately Marginally Not Total technologies suitable suitable suitable suitable Sand Storage Dam 3 5 2 0 10 Subsurface Dam 1 2 1 1 5 Total 4 7 3 1 15 Percentage 26.6 46.7 20 6.7 100 The validation results showed, Most of underground dam sites (46.7%) were found within the moderately suitable followed by highly suitable with 26.6%. The fact that most of predicted underground dam sits were found within marginally to highly suitable this indicates that the developed DSS tool can reliably be used to predict potential sites for underground dam sites. This method have more efficiency for locating suitable sites of sand storage dam than subsurface dam because process of selecting suitable site of subsurface dams are more complication and construct of that need more much cost. Altogether, the results showed for selecting underground dam sits can done economically based on this method with suitable accuracy, which in traditional method is done whit spending much money and long time in field work. Also, more than 50% of Iran is like as this study area, which could be use, this methodology for them too. 5. Conclusion This methodology describes a general method for the assessment of the suitability of sites for the installation of small underground dams (barrages). The selection criteria are defined both in a qualitative and quantitative way, and are based on a territorial analysis using satellite data and hydrological and climatologically information that are easily and freely available. The methodology is particularly indicated in areas where very little territorial information is available, such as central Iran, where the land planning is only based on large scale cartography which does not provide adequate information on environmental and morphological parameters. Altogether, the results showed for selecting underground dam sits can done economically based on this method with suitable accuracy, which in traditional method is done whit spending much money and long time in field work. Also, more than 50% of Iran is like as this study area, which could be use, this methodology for them too. Though the described approach requires a series of approximations, it is confirmed that the visual interpretation of satellite data and the analysis of large-scale availability cartography are precious instruments for an effective preliminary territorial analysis for the feasibility study of water resource projects in study area. It is important to observe that the analysis presented here determines the choice of suitable construction sites from prevalent engineering and technical perspectives. A conclusive evaluation about the feasibility and advantage of such projects Vol. 6, Issue 1, March 2015 26
should consider further perspectives. Many important socio-political aspects related to local and regional dynamics, which can influence the realization and effective benefit of the project, have been here neglected: transhumance route, dislocation of nomadic people, tensions between near villages, and other complex dynamics which are difficult to understand and quantify without a multidisciplinary approach. 6. Acknowledgment This paper was made possible thanks to the support of Agriculture and Natural Resource Research Center of Yazd and Department of Geography & Geology, Yerevan State University. We would like to thank the anonymous referees for providing positive comments. 7. References 1. Barkhordari J: The Calculation of potential Runoff and Selection of the Sites for Underground Dams in Yazd region of Iran (with use of RS, GIS and DSS), PhD thesis, Yerevan State University2013, P.151 2. Barkhordari J & Vardanian T: Using Post-classification Enhancement in Improving the Classification of Land use/cover of Arid Region (A case study in Pishkouh watershed, Center of Iran). Journal of rangeland science 2012, 2(2):p. 459-465 3. Barkhordari J & Semsar Yazdi AA: Assessment of the Monthly Water Balance in an Arid Region Using TM Model and GIS, Case Study: Pishkouh Watershed (Iran), Journal of rangeland science 2015,Vol.3,No.2(publishing) 4. Cavalcanti NB, Brito LT: Alternativa tecnologica para aumentar a disponibilidade de agua no semiarido. Revista Brasileira de Engenharia Agricola e Ambiental, Campina Grande 1999, 3 (1), 111 115 5. Cavalcanti N B, Resende G M: Avalicao de barreiros e finalidade da agua armazenada na regiao semi-arida da Bahia. Revista Brasileira de Engenharia Agricola e Ambiental, Campina Grande 2001,5 (3), 568 570 6. Forzieri G et.al: A methodology for the pre-selection of suitable sites for small surface dams in arid areas: A case study in the region of Kidal, Mali, Physics and Chemistry of the Earth Journal2008, No. 33,pp. 74 85, 7. Hansson G, Nilsson A: Groundwater dams for rural water supplies in developing countries. Groundwater 24 (4), 497 506, (1986). 8. FAO : Small dams and weirs in earth and gabion materials, Online press 2001 www.fao.org/docrep/012/i1531e/i1531e.pdf 9. Irura D: SubSurface Dams: a simple, safe and affordable technology for pastoralists. Report of TLDP Program in Kenia2006.50 pages. http://www.vsf-belgium.org 10. Mehrabi H R, Zeinivand H& Hadidi M: Site Selection for Groundwater Artificial Recharge in Silakhor Rangelands Using GIS Technique, Journal of rangeland science2012, Vol. 2, No. 4 pp.687-697 11. Nilsson A: Groundwater Dams for Small-scale Water Supply. Intermediate Technology Publications Ltd. 1988 12. Nilsson A: Sitting of groundwater dams for rural water supply in developing countries, Hydrogeologiacal and planning aspects. In: Proceedings of the Fifth World Congress on Water Resources1985, vol.3, pp. 1287 1296 13. Nissen Petersen E: Water from sand rivers, a manual on site survey, design,construction and maintenance of seven types of water structures in riverbeds, Sida s Regional Land Management Unit 2000, 57 pages. 14. Tabatabaii Yazdi J: Operation of subsurface flow from dry riverbeds using subsurface dams, SCWMRC. final research press2002, p.155 (in Farsi) Vol. 6, Issue 1, March 2015 27