Spatiotemporal variation of watershed health ...

Science of the Total Environment 587–588 (2017) 168–176

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Spatiotemporal variation of watershed health propensity through reliability-resilience-vulnerability based drought index (case study: Shazand Watershed in Iran) Seyed Hamidreza Sadeghi a,⁎, Zeinab Hazbavi b a b

Department of Watershed Management Engineering, Faculty of Natural Resources, Tarbiat Modares University, Iran Watershed Management Engineering, Faculty of Natural Resources, Tarbiat Modares University, Iran

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• SPI is a key watershed health assessment criterion. • Rel ResVul framework was conceptualized for SPI. • Spatiotemporal variations of SPI-Rel ResVul watershed health index was proved. • The weak response type of the sub-watersheds to climate condition was obtained.

a r t i c l e

i n f o

Article history: Received 27 October 2016 Received in revised form 25 January 2017 Accepted 11 February 2017 Available online 27 February 2017 Handling Editor: Wei Ouyang Keywords: Adaptation strategy Climate variability Ecosystem health assessment Environmental monitoring Land degradation Sustainable health

a b s t r a c t Quantitative response of the watershed health to climate variability is of critical importance for watershed managers. However, existing studies seldom considered the impact of climate variability on watershed health. The present study therefore aimed to analyze the temporal and spatial variability of reliability (Rel), resilience (Res) and vulnerability (Vul) indicators in node years of 1986, 1998, 2008 and 2014 in connection with Standardized Precipitation Index (SPI) for 24 sub-watersheds in the Shazand Watershed of Markazi Province in Iran. The analysis was based on rainfall variability as one of the main climatic drivers. To achieve the study purposes, the monthly rainfall time series of eight rain gauge stations distributed across the watershed or neighboring areas were analyzed and corresponding SPIs and Rel ResVul indicators were calculated. Ultimately, the spatial variation of SPI oriented Rel ResVul was mapped for the study watershed using Geographic Information System (GIS). The average and standard deviation of SPI-Rel ResVul index for the study years of 1986, 1998, 2008 and 2014 was obtained 0.240 ± 0.025, 0.290 ± 0.036, 0.077 ± 0.0280 and 0.241 ± 0.081, respectively. In overall, the results of the study proved the spatiotemporal variations of SPI-Rel ResVul watershed health index in the study area. Accordingly, all the sub-watersheds of the Shazand Watershed were grouped in unhealthy and very unhealthy conditions in all the study years. For 1986 and 1998 all the sub-watersheds were assessed in unhealthy status. Whilst, it declined to very unhealthy condition in 2008 and then some 75% of the watershed ultimately referred again to unhealthy and the rest still remained under very unhealthy conditions in 2014. © 2017 Elsevier B.V. All rights reserved.

⁎ Corresponding author. E-mail addresses: [email protected] (S.H. Sadeghi), [email protected] (Z. Hazbavi).

http://dx.doi.org/10.1016/j.scitotenv.2017.02.098 0048-9697/© 2017 Elsevier B.V. All rights reserved.

S.H. Sadeghi, Z. Hazbavi / Science of the Total Environment 587–588 (2017) 168–176

1. Introduction The global climate is changing (Kim and Chung, 2014). Change in rainfall as a driving force is one of the most critical factors determining the overall impact of climate change. However, rainfall is one of the foremost drivers of soil erosion by water. Hence, it is a major concern in soil conservation (Mengistu et al., 2016; Panagos et al., 2015; Sadeghi and Hazbavi, 2015; Brevik, 2013; Hazbavi and Sadeghi, 2013). Watersheds are dynamic biophysical landscape constructs that are driven by hydrological imperative (Falkenmark, 2003). The watershed conditions are strongly controlled by climate (long-term) and weather (short-term), hydrologic conditions, biotic/abiotic interactions and land uses. There is a belief among many scientists (e.g., Parkes et al., 2008 and Hazbavi and Sadeghi, 2016) that global climate changes will create “hydrological imperatives” and influence global precipitation patterns, altering both the amount of precipitation received and the distribution of precipitation (Brevik, 2013; IPCC, 2007) that require adaptation and management on a variety of scales. Effective watershed governance links health and sustainability with the concept of watersheds as ecosystems (Parkes et al., 2008). In recent years, due to the effect of climate change, drought studies are getting special attention (e.g., Li et al., 2016; Kumar et al., 2016; Paulo et al., 2016). Drought is characterized by below-average water availability (Kumar et al., 2016). It is often associated with large socioeconomic losses and damages to natural ecosystems leading to several environmental losses briefly are reflected in lowering quality and quantity of ecosystem services (Wilhite and Vanyarkho, 2000). On the other hand, rainfall anomalies in the more assessments of drought have characterized by Standardized Precipitation Index (SPI) as a probabilistic means of it (Keyantash and Dracup, 2002). The most recent report of the World Meteorological Organization (WMO) indicates that the SPI is the most prominent and primary meteorological index used (WMO, World Meteorological Organization, 2016). This index compares very favorably against several other “drought” indices (Giddings et al., 2005; Keyantash and Dracup, 2002). The SPI has been employed to examine numerous questions such as, drought, teleconnections with large scale circulatory systems, floods and crop yields (Giddings et al., 2005). McKee et al. (1993) developed the SPI to categorize observed rainfall as a standardized departure with respect to a rainfall probability distribution function. It indicates how precipitation for any given duration (e.g., one month, two months, etc.) at a particular observing site compares with the long-term precipitation record at the same site of the same duration (Edwards and McKee, 1997). Accordingly, different steps of time spans were applied to compute the SPI to facilitate the assessment of the effects of a precipitation deficit on different water resources components (soil moisture, groundwater, stream flow, reservoir storage) (Mashari Eshghabad et al., 2014). Several attempts have been made in the past to analyze the appropriateness in describing the drought characteristics for a particular region (e.g., Bandyopadhyay and Saha, 2016; Li et al., 2016; Kumar et al., 2016; Paulo et al., 2016; Xu et al., 2011; Jamshidi et al., 2011; Giddings et al., 2005; McKee et al., 1993) using the SPI. In addition, previous studies typically focused on the spatiotemporal trend analysis of the SPI (e.g., Wang et al., 2016; Rahmat et al., 2012; Tabari et al., 2011; Bonaccorso et al., 2003), future change analysis of the SPI (Bachmair et al., 2016; Rhee and Cho, 2016), investigating the relationship between the SPI and some remote sensing data (Damavandi et al., 2016; Owrangi et al., 2011). However, the SPI characterizing by risk assessment indicators viz. reliability (Rel), resilience (Res), and vulnerability (Vul), as the first attempt, has not been carried out or documented yet. The risk indicators including Rel ResVul can help watershed managers understand the behavior of the SPI changes and relate it to watershed health status. The watershed health may have several connotations. Here, the main focus was on the watershed health assessment based on important dynamic index of standardized precipitation and its spatiotemporal variation. The US EPA has proposed the use of integrated

169

assessments of watershed health to assist managers with identifying healthy watersheds and prioritizing candidate watersheds for protection and restoration (EPA, Environmental Protection Agency, 2014). A key component of watershed health is the ability to withstand, recover from, or adapt to natural and anthropogenic disturbances. However, healthy watersheds are naturally dynamic and often depend on recurrent natural disturbances to maintain their health (EPA, Environmental Protection Agency, 2012). Watershed health assessment (WHA) is therefore supposed as an appropriate conjunct approach between watershed research and management (Hazbavi and Sadeghi, 2016). Since a system failure should be simultaneously characterized by its Rel ResVul characteristics, a compound framework has to be used for comprehensive assessment of watershed health to a specific criterion like climatic variables. It is also a fundamental concept for developing new ways to assess and manage environmental resources (Suo et al., 2008). Application of Rel ResVul indicators in ecology (Walker et al., 2004; Naeem, 1998; Loucks, 1997; Holling, 1973) and hydrology (Asefa et al., 2014; Mondal et al., 2009; Kundzewicz and Laski, 1995; Hashimoto et al., 1982) was well documented. In addition, during last few years the Rel ResVul indicators have been employed by Sood and Ritter (2011) for watershed sustainability, Hoque et al. (2016, 2014, 2013, 2012) for water quality substances and by Chanda et al. (2014) and Maity et al. (2013) for Drought Management Index (DMI). Recently, a potential method of assessing watershed health with respect to hydrological responses was documented by Hazbavi and Sadeghi (2016). However, the application of Rel ResVul indicators coupled with the SPI has not been practically applied for the assessment of temporal and spatial variability of watershed health. Based on reviewing of the literature, it is very high important to investigate the watershed responses in respect to climate change. Since the climate change has the potential to disrupt and modify hydrological regimes and thus affects watershed health, it is very high essential to understand the response of the watershed system using this probabilistic framework in respect to SPI variation as a state-of-the-art method for assessing climatic variability. Hence, the present research attempted to expand the scope of the concept of Rel ResVul framework for the assessment of watershed health with the help of dynamic phenomenon of precipitation under governing changing climate. The results of the study reflect the general effectiveness of precipitation factor on changing watershed health. It, itself, helps managers and decision makers plan appropriately and adopt their adaptive approaches. Toward this attempt, the present research characterized the Rel ResVul indicators for the SPI and developed a SPIRel ResVul index for a semi-arid watershed located in central Iran for four node years of 1986, 1998, 2008 and 2014. 2. Materials and methods 2.1. Study area The present research was performed as a case study in the Shazand Watershed (ca. 1740 sq. km.) of south west of Markazi Province, Iran. It lies between the north latitudes 33° 44′42″ to 34° 12′ 13″ and east longitudes of 49° 04′ 15″ to 49° 52′ 12″ which is shown in Fig. 1. The Shazand Watershed is a part of Sareband mountainous City and located in a distance of 40 km from southwest of Arak City, capital of Markazi Province, and in the vicinity of Hamadan and Lorestan Provinces. The area underlying mainly by karst formation and therefore is important in terms of water resources (Yousefirad, 2005). Of total area, 50.15% includes highlands with hard formations and the rest 44.85% contains alluvial sediments and/or sub-mountain screes. According to Darabi et al. (2014), the Shazand Watershed has been divided into 24 sub-watersheds/inter-basins (Fig. 1). The mean annual precipitation in the watershed is about 420 mm, with a moderate semi-arid to cold semi-arid climate (Davudirad et al., 2016; Sadeghi and Hazbavi, 2016). During recent years, the Shazand Watershed experienced a huge development in

170


Fig. 1. General governing condition (right and upper) and geographical location of the Shazand Watershed in Iran with its sub-watersheds distribution (lower left).

agricultural activities, due to the potential for agricultural development along with industrial activities especially in the plain parts of the watershed. So, this development has exclusively increased overexploitation of ground water (Mokhtari et al., 2011; Davudirad et al., 2016). 2.2. SPI calculation For this research, the rainfall data was taken from the Ministry of Energy (http://www.moe.gov.ir/) and used for the calculation of SPI as developed by McKee et al. (1993) to categorize observed rainfall as a standardized departure with respect to a rainfall probability distribution function. Accordingly, the data from eight rain gauge stations viz. Arak, Astaneh, Emarat, Ghadamgah, Khondab, Mazrae Khatun, Gavar and Sarugh were collected. The distribution of the study rain gauge stations and summary of their characteristics are presented in Fig. 1 and Table 1, respectively. A frequency distribution of precipitation data was built to determine the SPI (Sigdel and Ikeda, 2010) at the study stations for the period of 1977–2014. It indicates how precipitation for any given duration (here one month), at a particular observing site compares with the longterm precipitation record at the same site of the same duration (Edwards and McKee, 1997). The SPI values have been calculated at

monthly time scale using Drought Index Package (DIP) software package (Morid et al., 2005). 2.3. Threshold selection of the SPI Since the absolute thresholds of the evaluation indicators for the regional ecosystem health do not exist, the SPI values for three months reported by WMO, World Meteorological Organization (2016) was applied for the present research. Aside from identifying indicators important for drought impacts, there is a need for a better understanding of the meaning of indicator thresholds used for drought declaration and as triggers for management actions in drought plans (Bachmair et al., 2016; Nadal-Romero et al., 2016). Such thresholds are mostly based on hazard intensity classes corresponding to a certain frequency of occurrence. For instance, following the widely accepted SPI scheme, with classes ranging from zero to −0.99 (mild drought), −1 to −1.49 (moderate drought), −1.5 to −2 (severe drought), and less than − 2 (extreme drought) (McKee et al., 1993). By identifying the SPI as an index for broad use, WMO, World Meteorological Organization (2016) provided direction for countries trying to establish a level of drought early warning. Drought events are indicated when the results of SPI, for whichever timescale is being investigated, become continuously

S.H. Sadeghi, Z. Hazbavi / Science of the Total Environment 587–588 (2017) 168–176 Table 1 Rain gauge stations used for analysis of SPI-Rel ResVul index for different years in the Shazand Watershed, Iran. Station

X

Y

Year

Rainfall

SPI

Arak

380,457

3,772,091

1986 1998 2008 2014 1986 1998 2008 2014 1986 1998 2008 2014 1986 1998 2008 2014 1986 1998 2008 2014 1986 1998 2008 2014 1986 1998 2008 2014 1986 1998 2008 2014

358.10 281.30 190.80 336.80 587.00 461.50 362.50 489.50 462.00 326.00 265.50 388.20 382.00 456.30 257.50 272.50 589.50 490.50 669.50 395.00 362.00 180.00 363.00 216.00 397.50 265.00 378.00 298.00 316.00 181.50 303.00 180.00

−0.03 0.21 −0.40 −0.01 0.06 0.00 −0.19 0.00 −0.06 −0.11 −0.38 0.02 −0.12 0.27 −0.29 −0.30 0.04 0.23 0.43 −0.18 −0.01 −0.35 0.30 −0.17 0.03 −0.20 0.18 −0.12 0.00 0.01 0.22 −0.30

Astaneh

348,056

3,750,772

Emarat

368,607

3,748,287

Ghadamgah

358,697

3,760,305

Mazrae Khatun

330,140

3,762,193

Khondab

333,452

3,805,795

Gavar

Sarugh

377,673

361,441

3,759,505

3,808,925

negative and reach a value of −1. The drought event is considered to be ongoing until SPI reaches a value of zero. Accordingly, McKee et al. (1993) stated that the drought begins at a SPI ≤ −1. But there is no standard in place, as some researchers will choose a threshold that is less than zero, but not quite −1, whilst others will initially classify drought at values less than −1. For the present research, with examining different thresholds such as min, max, mean, median and other criteria, it was concluded that the threshold of 0.1 was the best to show the changes of the Rel ResVul indicators for the present research. The threshold level of 0.1 for SPI was selected to not only show the likely occurrence of normal conditions in the study area but also be able to rectify the variability of the SPI during the study period.

2.4. Rel ResVul indicators calculation Three risk indicators, namely Rel ResVul are calculated in the context of drought through the analysis of spatial and temporal variation of the SPI. These indicators measure changes in the status of a system or sub-system over which an organization or several organizations have responsibility, e.g., a watershed. The simple R el ResVul framework was developed as a common framework for the context of water resources management. In this context, the Rel (frequency), R es (ability to recover) and V ul of each rain gauge station of the Shazand Watershed from a SPI as drought index assessment were computed based on the approach was given by Hashimoto et al. (1982) and Kjeldsen and Rosbjer (2004). In the Rel ResVul framework, the Rel is defined by the probability that a watershed is in a satisfactory state (Hoque et al., 2016, 2012; Hashimoto et al., 1982) in monthly time period (Jain, 2010) in view point of the SPI. In the context of drought, the Rel ResVul is defined as the probability that the SPI is beyond a certain threshold (here 0.1). Accordingly, the watershed Rel to SPI was calculated as described in Eq.

171

(1). M

Rel ¼ 1−∑ j¼1 dð jÞ

. T

ð1Þ

where d(j) is the duration of the jth failure event, M is the number of failure events, and T is the total number of time intervals (here 12). Moreover, Res is a measure that indicates how quickly the watershed can return to a satisfactory stage after it has fallen below the satisfactory threshold. This can be defined as the ratio of the probability of transition from the unsatisfactory to the satisfactory stage and the probability of failure (Maity et al., 2013). The Res was determined from Eq. (2). ( Res ¼

)−1 1 M ∑ dð jÞ M j¼1

ð2Þ

The last indicator of the Rel ResVul framework, the Vul as a measure of the likely damage of a failure event was computed for each time using Eq. (3).

Vul ¼

. 1 T ∑ Lobs ðiÞ−Lstd ðiÞ H½Lobs ðiÞ−Lstd ðiÞ Lstd ðiÞ M i¼1

ð3Þ

where Lobs(i) is the observed study constituent at the ith time step, Lstd(i) is the corresponding compliance standard (here 0.1), and H( ) is the Heaviside function which ensures that only failure events are involved in the Vul calculation in Eq. (3). In the present research the Rel ResVul indicators are unitless. The Rel and Res are measured on the probability scale, i.e., 0–1. Maity et al. (2013) found that Rel and Res follow a similar trend. The effect of climate change stressor on the watershed health depends on the vulnerability of the watershed and its ability to adapt to the change. Changes in precipitation characteristics and mainly regimes as indicated by SPI variation affect many ecological processes having a natural range of variability and ultimately lead to general status of the watershed heath. So that, the watershed conditions either promote to healthier situation or resist against climate change or getting retrogressive trend and need to be assist (EPA, Environmental Protection Agency, 2012). 2.5. Rel ResVul indicators standardization Among the selected indicators, a subset (Vul) has a negative influence on SPI-Rel ResVul index, whereas others (Rel and Res) have a positive effect. Standardization adjusts or controls for differences in Rel ResVul indicators and provides a single summary measure for the comparison. The resulting adjusted value is an artificial rate that allows for comparisons over time and place. The standardization can be used to adjust for any one underlying factor. The indicators are standardized to range from zero to one using the description of Eq. (4). Y¼

xi −xmin xmax −xmin

ð4Þ

where, Y is the standardized value of each indicator included in the assessment; xi is the actual value; and xmax and xmin are the maximum and the minimum values observed data in all actual measurements, respectively (Zhao et al., 2006; Hazbavi and Sadeghi, 2016). 2.6. SPI-Rel ResVul index calculation The three indicators of Rel ResVul were calculated for each station, individually, in the study watershed. Then the results were standardized via the explanation technique (Wiegand et al., 2013; Loucks, 1997).

172


The SPI-Rel ResVul index can thus be expressed as Eq. 5. 1

SPI−Rel Res Vul index ¼ ðRel Res Vul Þ3

ð5Þ

Here, this method is known as the Rel ResVul framework and is also referred to as the SPI-Rel ResVul index. 2.7. SPI-Rel ResVul index classification The relative value estimation of the individual and comprehensive indices is used to ensure the consistency of the assessment criterion among indicators. After ranking the Rel ResVul values from high to low, the regional SPI-Rel ResVul index state is divided into five degrees (Hazbavi and Sadeghi, 2016; Yu et al., 2013; Sang et al., 2006), which reflect the SPI-Rel ResVul index state change from “very healthy” to “very unhealthy” as described in Table 2. 2.8. SPI-Rel ResVul index mapping Based on the previous productions, the Rel ResVul indicators and SPIRel ResVul index were calculated for each rain gauge station. The SPI-Rel ResVul index distribution subsequently was mapped using the function of spatial expression in Geographic Information System (GIS) to monitor and assess ecosystem conditions and changes on multiple scales (Kolios and Stylios, 2013; Yu et al., 2013). The data were interpolated using the Inverse Distance Weighted (IDW) method (Sadeghi et al., 2016; Millington et al., 2015; Fathizad et al., 2014) in the ArcGIS 10.3. The IDW is an algorithm for spatially interpolating or estimating values between measurements. Each value estimated in an IDW interpolation is a weighted average of the surrounding representative points. Weights are computed by taking the inverse of the distance from an observation location to the location of the point being estimated (Prasannakumar et al., 2011; Burrough and McDonnell, 1998) and these yield more consistent result than other methods with original input data.

Fig. 2. The results of annual rainfall and SPI during 1977–2014 period for the Shazand Watershed, Iran.

3. Results and discussion Time series of annual rainfall and calculated SPI were presented in Fig. 2. The mean rainfall for node years of 1986, 1998, 2008 and 2014 were 455.73, 398.28, 234 and 347 mm, respectively. The corresponding SPI values were also obtained 0.97, −0.01, −1.12 and −3.18 for 1986, 1998, 2008 and 2014, respectively. As shown in Fig. 2, the minimum rainfall and the SPI were found for 2008 and 2014, respectively. It can be concluded that the there is no distinct decreasing trend in rainfall whereas a significant declining trend is seen in the SPI verifying increasing tendency in drought severity in the Shazand Watershed. It is assumed that the watershed behavior into a SPI-Rel ResVul index might have different patterns. The results of Rel ResVul indicators in studied sub-watersheds and years have been presented in Table 3. The minimum SPI Rel ResVul values of 0.35, 0.26 and 10.18 were obtained for 2008 and the maximum SPI Rel ResVul values were calculated 0.65, 0.74 and 33.84 for 1986, 0.72, 0.49 and 74.16 for 1998, 0.56, 0.69 and 25.04 for 2008 as well as 0.74, 0.96 and 26.70 for 2014. In the next step, the Rel ResVul indicators were standardized and SPIRel ResVul index was computed for all the sub-watersheds/inter-basin and study years, consequently. The spatiotemporal variations of SPIRel ResVul index across the Shazand Watershed during 1986, 1998, 2008 and 2014 were presented in Figs. 3 and 4.

The condition of SPI-Rel ResVul index of whole the Shazand Watershed is classified in two categories of unhealthy and very unhealthy. The minimum SPI-Rel ResVul index was obtained for 2008 with 0.077 ± 0.0280 and classified in unhealthy category. SPI-Rel ResVul index was calculated for 1986, 1998 and 2014 with 0.240 ± 0.025, 0.290 ± 0.036 and 0.241 ± 0.081, respectively was classified in very unhealthy category. The results for 1986 year in spatial scale showed that north, northwest and west sub-watersheds of the study area had unhealthy SPI-Rel ResVul index status. In addition, for this year, south and southeast subwatersheds had the best SPI-Rel ResVul index status. From Fig. 4, spatial variation in the 1998 year is clearly visible. In the 1998 year, the unhealthy status of SPI-Rel ResVul index from the south and southeast sub-watersheds shifted to north, northwest and west sub-watersheds of the study area. However, for these two years the all of sub-watersheds in view point of SPI-Rel ResVul index are categorized in the unhealthy status (0.21–0.40). The Rel ResVul analysis also revealed that the all sub-watersheds were in the very unhealthy status (0.00–0.20) in view point of SPI-Rel ResVul index for 2008 year. However, the sub-watersheds of two, 10, 12 and 13 had the better SPI-Rel ResVul index rather than others. In addition, as shown in Fig. 3 SPI-Rel ResVul index in the all sub-watersheds in the year of 2008 was less than other study years. This indicated that the

Table 2 Description of classification for SPI-Rel ResVul index for watershed health assessment. Category

I

II

III

IV

V

Values Description

0.81–1.00 Very healthy

0.61–0.80 Healthy

0.41–0.60 Moderately healthy

0.21–0.40 Un-healthy

0.00–0.20 Very un-healthy


173

Table 3 Summary of Rel ResVul indicators of SPI for four years of 1386, 1998, 2008 and 2014 in the different sub-watersheds. Year

1986

1998

2008

2014

Indicator Sub-watershed 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

R

R

V

R

R

V

R

R

V

R

R

V

0.645 0.531 0.536 0.559 0.509 0.621 0.516 0.505 0.549 0.552 0.578 0.516 0.514 0.611 0.509 0.533 0.648 0.520 0.617 0.600 0.580 0.569 0.541 0.548

0.724 0.629 0.545 0.506 0.505 0.699 0.546 0.507 0.514 0.634 0.512 0.622 0.649 0.700 0.511 0.495 0.742 0.526 0.678 0.623 0.475 0.492 0.566 0.523

22.792 28.158 30.499 23.775 29.084 24.942 28.757 28.237 31.387 27.825 32.380 28.262 27.620 25.784 28.677 33.780 22.593 29.526 25.097 28.035 33.839 33.434 30.518 32.118

0.587 0.683 0.629 0.568 0.663 0.592 0.665 0.664 0.540 0.653 0.500 0.700 0.718 0.594 0.663 0.662 0.588 0.653 0.591 0.564 0.479 0.518 0.619 0.604

0.412 0.362 0.443 0.462 0.488 0.428 0.452 0.491 0.459 0.395 0.474 0.377 0.354 0.442 0.485 0.287 0.417 0.466 0.414 0.437 0.475 0.469 0.415 0.456

28.238 52.809 56.408 56.704 48.015 34.405 50.017 45.930 49.096 50.396 63.915 54.325 54.038 35.453 47.364 60.084 27.211 51.039 38.080 49.832 74.157 67.472 61.019 59.380

0.555 0.476 0.413 0.424 0.353 0.525 0.393 0.346 0.422 0.482 0.401 0.467 0.489 0.513 0.355 0.394 0.558 0.381 0.520 0.473 0.379 0.394 0.436 0.409

0.696 0.362 0.313 0.332 0.265 0.575 0.294 0.260 0.334 0.406 0.324 0.332 0.345 0.506 0.267 0.287 0.686 0.287 0.588 0.466 0.304 0.310 0.325 0.315

13.638 21.857 19.042 20.002 11.316 15.710 15.059 10.180 18.526 20.876 22.587 22.161 23.580 16.501 11.299 17.740 13.707 14.560 16.485 19.718 25.040 23.058 22.595 19.794

0.664 0.712 0.657 0.683 0.539 0.656 0.599 0.522 0.683 0.704 0.702 0.711 0.736 0.640 0.539 0.635 0.653 0.589 0.680 0.688 0.729 0.708 0.715 0.666

0.937 0.894 0.688 0.764 0.280 0.849 0.496 0.237 0.795 0.900 0.876 0.877 0.965 0.769 0.289 0.515 0.904 0.455 0.931 0.900 0.953 0.871 0.894 0.707

27.120 19.586 22.142 22.725 23.371 24.756 21.831 22.887 20.176 20.908 22.606 18.836 17.833 22.980 22.956 26.124 26.702 22.692 25.321 23.822 23.671 23.215 20.827 23.260

status of the Rel and Res in this year was very decreased and the watershed didn't can recover its very well against rainfall decreases (Fig. 2) and consequently the Shazand Watershed Vul was increased rather than the study years. For 2014, the SPI-Rel ResVul index got a better status rather than 2008. In the year of 2014, sub-watersheds five, seven, eight, 15, 16 and 18 had the SPI-Rel ResVul index between 0.00 and 0.21 and had very unhealthy status. The other sub-watersheds were classified in unhealthy status with values of SPI-Rel ResVul index between 0.21 and 0.40. As noted by Falkenmark (2003), a low resilience of the watershed through drought may cause conflict over critical resources such as freshwater and food leading to loss of livelihood and trigger tension. It clearly emphasizes the necessity of adaptation of managerial approaches for sustainable successful management of soil and water resources. Thus, the appropriate perception on intensifying SPI provides managers some advance information to expect over the next years at any given sub-watersheds accordingly adjust their managerial measures and approaches. A similar emphasis has been made by Chanda et al. (2014) in regards to drought management index. Suo et al. (2008) also found that the unhealthy sub-watersheds, the Upper Hua-njiang, China, scored the lowest (0.296) ecosystem health indicators. They related this weak response type of the sub-watersheds to its climate conditionbecause, most of the northern portion of the Huangtu Plateau laid in a semi-arid climate similar to Shazand Watershed. In addition, the obtained results were verified by Damavandi et al. (2016) who showed that Markazi Province including the Shazand Watershed in the period of 2000–2014 had a moderate and low drought classes but had an unhealthy condition in view point of land degradation. Based on the finding of Hoque et al. (2013) climate change scenarios that involved rising precipitation levels were found to negatively impact Rel ResVul indicators with respect to sediment, total N and total P load. The same Rel ResVul indicators variation was verified in the previous researches which conducted for different criteria. In this regards, Chanda et al. (2014) who applied the Rel ResVul indicators for assessing soil moisture data in India as well as status of DMI over the past 50 years (1961–1965 to 2006–2010), indicated that drought propensity is consistently low toward north and northeast parts of India, but much higher in the west part of peninsular India. In this regards and the same region, DMI (Maity et al., 2013) is designed such that it increased with increase in vulnerability as well as with decrease in resilience and vice versa.

The Rel ResVul analysis conducted by Hoque et al. (2012) under uncertainty revealed that the Cedar Creek watershed health located in Indiana, USA is in overall good condition for atrazine, alachlor, ammonia and total phosphorus (P) based on water quality data. However, the watershed was found to be susceptible to sediment violations. Hoque et al. (2014) examined the spatial nature of Rel ResVul indicators with respect to the same water quality constituents (sediment, total N and total P) in the five agricultural watersheds in the USA Midwest. Scaling Rel ResVul indicators was examined both in terms of contributing area and Strahler number. Large fluctuations were observed in Rel ResVul values at smaller drainage areas for both sediments and nutrients, before converging toward stable quantities. They noted that the variance in corresponding Rel ResVul values at lower stream orders would be high. This spread would gradually reduce and eventually converge to a stable value at outlets of higher order streams. In addition, Hoque et al. (2016) found that individual and aggregate Rel ResVul indicators for water quality including atrazine, alachlor, NH4-N, total P, total suspended solids (TSS), NO3-NO2-N and total Kjeldahl N, were very varied from place to place conducted on two watersheds of Indiana, USA. They concluded the Rel ResVul values is very usefulness as a tool to identify area of particular concern within a watershed. Although the methodology was based on the measured data and real conditions of the study watershed with high level of precision, but no realistic or tangible criterion could be used for validation of the outcome of the approach. However, anecdotal evidences on general decreasing tendency in the watershed services representing in water shortages, distinct variations in the relationship between streamflow and sediment and also changes within different periods, as clearly reported by Davudirad et al. (2016) for the same watershed, prove the soundness of the results on the assessment of watershed general health. 4. Conclusions With the consequences of climate change as it is postulated the droughts would become more common in the future. Along with this increased vulnerability, concern exists because some research suggests that drought in the future may be amplified in certain areas due to changes in climate variability and extremes resulting from global warming (www.ncdc.noaa.gov). The indices of Rel ResVul in different time and space scales can be therefore be used to measure different aspects of number, extent, and severity of the performance of a watershed in respect to rainfall anomaly. According to the Rel ResVul framework, the

174


Fig. 3. The results of SPI-Rel ResVul index for different Sahzand sub-watersheds and years.

scores from three indicators (Rel ResVul) and a comprehensive assessment index (SPI-Rel ResVul index) were calculated and an assessment map was constructed. According to results the present research, non-significant and significant decreasing trend was respectively found for rainfall and SPI which verified increasing tendency in drought severity in the Shazand

Watershed. The main finding showed that the SPI-Rel ResVul index of 0.240 ± 0.025, 0.290 ± 0.036, 0.077 ± 0.0280 and 0.241 ± 0.081, respectively, was obtained for 1986, 1998, 2008 and 2014. Based on the Rel ResVul analyses, it may be concluded that the Shazand Watershed health was in overall unhealthy condition in viewpoint of SPI for the study years. Spatiotemporal variation of SPI-Rel ResVul index was


175

Fig. 4. The spatial arrangement of the SPI-Rel ResVul index in the Shazand Watershed for 1986, 1998, 2008 and 2014 years.

confirmed that the north, northwest and west of the Shazand Watershed experienced the lowest SPI-Rel ResVul index in 1986. Then the minimum level of SPI-Rel ResVul was then shifted to north, northwest and west of the watershed in 1998. In addition, the variation of spatial distribution in the SPI-Rel ResVul index was clearly observed in 2008 and 2014 for Shazand Watershed. A focus on maintaining and enhancing watershed resilience will enable better adaptation to existing and future climate and weather events. The provincial government could therefore provide ecosystem management and planning as a climate mitigation/adaptation strategy. However, further investigations are needed to assess the sensitivity of the Rel ResVul framework to different SPI thresholds. Evaluation of compound effects of other criteria associated with human activities and climatic factors on changing health of the Shazand Watershed by Rel ResVul framework is also essentially needed. Such studies serve to expand the more knowledge base on watershed conditions to different driving forces and will inform future efforts to better improve watershed degradation controls and treatments at various spatial and temporal scales. Though more insight and elaborated studies considering complementary indices are needed for comprehensive assessment of the watershed health.

Acknowledgments The authors would like to thank Eng. Fatemeh Ghanbari, Technical Expert of Department of Watershed Management Laboratory and Eng. Davudirad, Ph.D. Student of Tarbiat Modares University for their assistances in data collection and documentation. The authors also are grateful to anonymous reviewers and respected associate editor of the Science of the Total Environment Journal, Dr. Wei Ouyang, for their

valuable comments and suggestions that substantially improved the manuscript. Appendix A. Supplementary data Supplementary data associated with this article can be found in the online version, at doi: 10.1016/j.scitotenv.2017.02.098. These data include the Google maps of the most important areas described in this article. References Asefa, T., Clayton, J., Adams, A., Anderson, D., 2014. Performance evaluation of a water resources system under varying climatic conditions: reliability, resilience, vulnerability and beyond. J. Hydrol. 508, 53–65. Bachmair, S., Svensson, C., Hannaford, J., Barker, L.J., Stah, K., 2016. A quantitative analysis to objectively appraise drought indicators and model drought impacts. Hydrol. Earth Syst. Sci. 20, 2589–2609. Bandyopadhyay, N., Saha, A.K., 2016. A comparative analysis of four drought indices using geospatial data in Gujarat, India. Arab. J. Geosci. 9 (341) (11 pp). Bonaccorso, B., Bordi, I., Cancielliere, A., Rossi, G., Sutera, A., 2003. Spatial variability of drought: an analysis of the SPI in Sicily. Water Resour. Manag. 17, 273–296. Brevik, E.C., 2013. The potential impact of climate change on soil properties and processes and corresponding influence on food security. Agriculture 3, 398–417. Burrough, P.A., McDonnell, R.A., 1998. Principles of Geographical Information Systems for Land Resources Assessment. J. Quat. Sci. 3(1). Oxford University Press, New York (108 pp). Chanda, K., Maity, R., Sharma, A., Mehrotra, R., 2014. Spatiotemporal variation of longterm drought propensity through reliability-resilience-vulnerability based drought management index. Water Resour. Res. 50, 7662–7676. Damavandi, A.A., Rahimi, M., Yazdani, M.R., 2016. Assessment of drought severity using vegetation temperature condition index (VTCI) and Terra/MODIS satellite data in rangelands of Markazi Province, Iran. J. Rangeland Sci. 6 (1), 33–42 (In Persian). Darabi, H., Shahedi, K., Solaimani, K., Miryaghoubzadeh, M., 2014. Prioritization of subwatersheds based on flooding conditions using hydrological model, multivariate analysis and remote sensing technique. Water Environ. J. 28, 382–392.

176


Davudirad, A.A., Sadeghi, S.H.R., Sadoddin, A., 2016. The impact of development plans on hydrological changes in the Shazand Watershed. Iran Land Degrad. Dev. 27, 1236–1244. Edwards, D.C., McKee, T.B., 1997. Characteristics of 20th century drought in the United States at multiple time scales. Climatology Report Number 97–2. Colorado State University, Fort Collins, Colorado 174 pp. EPA (Environmental Protection Agency), 2012. Identifying and Protecting Healthy Watersheds. 34 pp. available at: http://water.epa.gov/healthywatersheds. EPA (Environmental Protection Agency), 2014. Wisconsin Integrated Assessment of Watershed Health. 111 pp. available at: http://www.epa.gov/healthywatersheds. Falkenmark, M., 2003. Water Management and Ecosystems: Living with Change. Global Water Partnership Technical Advisory Committee Background Paper No. 9, Stockholm (44 pp). Fathizad, H., Karimi, H., Alibakhsh, S.M., 2014. The estimation of erosion and sediment by using the RUSLE model and RS and GIS techniques (Case study: arid and semi-arid regions of Doviraj, Ilam province, Iran). Intl. J. Agri. Crop Sci. 7 (6), 304–314. Giddings, L., Soto, M., Rutherford, B.M., Maarouf, A., 2005. Standardized precipitation index zones for México. Atmosfera 2015, 33–56. Hashimoto, T., Loucks, D.P., Stedinger, J., 1982. Reliability, resilience and vulnerability for water resources system performance evaluation. Water Resour. Res. 18, 14–20. Hazbavi, Z., Sadeghi, S.H.R., 2013. Analysis of spatial trend of rainfall erosivity in Iran. First International Conference on Environmental Crisis and Its Solutions, Kish Island-Iran, pp. 434–440. Hazbavi, Z., Sadeghi, S.H.R., 2016. Watershed health characterization using reliability-resilience-vulnerability conceptual framework based on hydrological responses. Land Degrad. Dev. http://dx.doi.org/10.1002/ldr.2680. Holling, C.S., 1973. Resilience and stability of ecological systems. Annu. Rev. Ecol. Evol. Syst. 4, 1–23. Hoque, Y.M., Tripathi, S., Hantush, M.M., Govindaraju, R.S., 2012. Watershed reliability, resilience and vulnerability analysis under uncertainty using water quality data. J. Environ. Manag. 10, 101–112. Hoque, Y.M., Raj, C., Hantush, M.M., Chaubey, I., Govindaraju, R.S., 2013. How do land-use and climate change affect watershed health? A scenario-based analysis. Water Qual Expo Health 6, 19–33. Hoque, Y.M., Hantush, M.M., Govindaraju, R.S., 2014. On the scaling behavior of reliability–resilience–vulnerability indices in agricultural watersheds. Ecol. Indic. 40, 136–146. Hoque, Y.M., Tripathi, S., Hantush, M.M., Govindaraju, R.S., 2016. Aggregate measures of watershed health from reconstructed water quality data with uncertainty. J. Environ. Qual. 45, 709–719. IPCC, 2007. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L. (Eds.), Summary for Policymakers. In Climate Change 2007: The Physical Science Basis; Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK, pp. 1–18. Jain, S.K., 2010. Investigating the behavior of statistical indices for performance assessment of a reservoir. J. Hydrol. 391, 90–96. Jamshidi, H., Khalili, D., Zadeh, M.R., Hosseinipour, E.Z., 2011. Assessment and comparison of SPI and RDI meteorological drought indices in selected synoptic stations of Iran. World Environmental and Water Resources Congress. 2011, pp. 1161–1173. Keyantash, J., Dracup, J.A., 2002. The quantification of drought: an evaluation of drought indices. Bull. Am. Meteorol. Soc. 83 (8), 1167–1180. Kim, Y., Chung, E.S., 2014. An index-based robust decision making framework for watershed management in a changing climate. Sci. Total Environ. 473-474, 88–102. Kjeldsen, T.R., Rosbjer, D., 2004. Choice of reliability, resilience and vulnerability estimators for risk assessments of water resources systems. Hydrol. Sci. J. 49, 755–767. Kolios, S., Stylios, C.D., 2013. Identification of land cover/land use changes in the greater area of the Preveza peninsula in Greece using Landsat satellite data. Appl. Geogr. 40, 150–160. Kumar, R., Musuuza, J.L., Van Loon, A.F., Teuling, A.J., Barthel, R., Broek, J.T., Mai, J., Samaniego, L., Attinger, L., 2016. Multiscale evaluation of the standardized precipitation index as a groundwater drought indicator. Hydrol. Earth Syst. Sci. 20, 1117–1131. Kundzewicz, Z.W., Laski, A., 1995. Reliability-related criteria in water supply studies. New Uncertainties Concepts in Hydrology and Water Resources. Cambridge University Press, Cambridge, UK, pp. 299–305. Li, J., Zhou, S., Hu, R., 2016. Hydrological drought class transition using SPI and SRI time series by Log linear regression. Water Resour. Manag. 30, 669. Loucks, D.P., 1997. Quantifying trends in system sustainability. Hydrol. Sci. J. 42 (4), 513–530. Maity, R., Sharma, A., Nagesh Kumar, D., Chanda, K., 2013. Characterizing drought using the reliability-resilience-vulnerability concept. J. Hydraul. Res. 18, 859–869. Mashari Eshghabad, M., Omidvar, E., Solaimani, K., 2014. Efficiency of some meteorological drought indices in different time scales (case study: Tajan Basin, Iran). ECOPERSIA 2 (1), 441–453. McKee, T.B.N., Doesken, J., Kliest, J., 1993. The relationship of drought frequency and duration to time scales. 8th Conference on Applied Climatology, Anaheim, CA, American Meteorological Society, Washington DC, pp. 179–184. Mengistu, D., Bewket, W., Lal, R., 2016. Conservation effects on soil quality and climate change adaptability of Ethiopian Watersheds. Land Degrad. Dev. 27 (6), 1603–1621. Millington, H.K., Lovell, J.E., Lovell, C.A.K., 2015. A framework for guiding the management of urban stream health. Ecol. Econ. 109, 222–233. Mokhtari, Sh., Babazadeh, H., Sedghi, H., Kaveh, F., 2011. Long term simulation of Shazand Plain Aquifer under changing resources and applications. Int. J. Agr. Sci. Res. 2, 1–10.

Mondal, M.S., Chowdhury, J.U., Ferdous, M.R., 2009. Risk based evaluation for meeting future water demand of the Brahmaputra Floodplain within Bangladesh. Water Resour. Manag. 24 (5), 853–869. Morid, S., Moghaddasi, M., Arshad, S., Omid, M., 2005. Drought Index Package (Version 2). Tarbiat Modares University, Tehran, Iran. Nadal-Romero, E., Lasanta, T., Cerdà, A., 2016. Integrating extensive livestock and soil conservation policies in Mediterranean Mountain areas for recovery of abandoned lands in the Central Spanish Pyrenees, a long-term research assessment. Land Degrad. Dev. 2016, 1–12. Naeem, S., 1998. Species redundancy and ecosystem reliability. Conserv. Biol. 12 (1), 39–45. Owrangi, M.A., Adamowski, J., Rahnemaei, M., Mohammadzadeh, A., Afshin Sharifan, R., 2011. Drought monitoring methodology based on AVHRR images and SPOT vegetation maps. J. Water Resour. Prot. 3, 325–334. Panagos, P., Ballabio, C., Borrelli, P., Meusburger, K., Klik, A., Rousseva, S., Tadić, M.P., Michaelides, S., Hrabalíková, M., Olsen, P., Aalto, J., Lakatos, M., Rymszewicz, A., Dumitrescu, A., Beguería, S., Alewell, C., 2015. Rainfall erosivity in Europe. Sci. Total Environ. 511, 801–814. Parkes, M.W., Morrison, K.E., Bunch, M.J., Venema, H.D., 2008. Ecohealth and watersheds: ecosystem approaches to re-integrate water resources management with health and well-being. Network for Ecosystem Sustainability and Health (Publication Series No. 2) and the International Institute for Sustainable Development, Winnipeg, MB Available online at. http://www.iisd.org/pdf/2008/ecohealth_watersheds.pdf (74 pp). Paulo, A., Martins, D., Pereira, L.S., 2016. Influence of precipitation changes on the SPI and related drought severity. An analysis using long-term data series. Water Resour. Manag. 2016, 1–21. Prasannakumar, V., Vijith, H., Geetha, N., Shiny, R., 2011. Regional scale erosion assessment of a sub-tropical highland segment in the Western Ghats of Kerala, South India. Water Resour. Manag. 25, 3715–3727. Rahmat, S.N., Jayasuriya, N., Bhuiyan, M., 2012. Trend analysis of drought using standardised precipitation index (SPI) in Victoria, Australia. 34th Hydrology and Water Resources Symposium, 19–22 November 2012, Sydney, Australia, pp. 441–448. Rhee, J., Cho, J., 2016. Future changes in drought characteristics: regional analysis for South Korea under CMIP5 projections. Bull. Am. Meteorol. Soc. 17, 437–451. Sadeghi, S.H.R., Hazbavi, Z., 2015. Trend analysis of the rainfall erosivity index at different time scales in Iran. Nat. Hazards 77, 383–404. Sadeghi, S.H.R., Hazbavi, Z., 2016. Watershed health assessment based on soil loss using reliability, resilience and vulnerability framework. 19th PSSST Annual Meeting and Scientific Conference, Legazpi City, Albay, 18–20 May (1 p). Sadeghi, S.H.R., Zabihi, M., Vafakhah, M., Hazbavi, Z., 2016. Spatiotemporal mapping of rainfall erosivity index for different return periods in Iran. Nat. Hazards http://dx. doi.org/10.1007/s11069-017-2752-3. Sang, Y., Chen, X., Wu, R., Peng, X., 2006. Comprehensive assessment of urban ecosystem health. Chin. J. Appl. Ecol. 17 (7), 1280–1285. Sigdel, M., Ikeda, M., 2010. Spatial and temporal analysis of drought in Nepal using standardized precipitation index and its relationship with climate indices. J. Hydrol. Meteorol. 1, 59–74. Sood, A., Ritter, W.F., 2011. Developing a framework to measure watershed sustainability by using hydrological/water quality model. J. Water Resour. Prot. 3, 788–804. Suo, A.N., Xiong, Y.C., Wang, T.M., Yue, D.X., Ge, J.P., 2008. Ecosystem health assessment of the Jinghe River Watershed on the Huangtu Plateau. EcoHealth 5, 127–136. Tabari, H., Marofi, S., Aeini, A., Talaee, P.H., Mohammadi, K., 2011. Trend analysis of reference evapotranspiration in the western half of Iran. Agric. For. Meteorol. 151, 128–136. Walker, B., Holling, C.S., Carpenter, S.R., Kinzig, A., 2004. Resilience, adaptability and transformability in social-ecological systems. Ecol. Soc. 9 (2) 5 pp. Wang, Y., Xu, Y., Lei, C., Li, G., Han, L., Song, S., Yang, L., Deng, X., 2016. Spatio-temporal characteristics of precipitation and dryness/wetness in Yangtze River Delta, eastern China, during 1960–2012. Atmos. Res. 172-173, 196–205. Wiegand, A.N., Walker, C., Duncan, P.F., Roiko, A., Tindale, N., 2013. A systematic approach for modelling quantitative lake ecosystem data to facilitate proactive urban lake management. Environ. Syst. Res. 2 12 pp. Wilhite, D.A., Vanyarkho, O.V., 2000. Drought: Pervasive Impacts of a Creeping Phenomenon. Drought Mitigation Center Faculty Publications Paper 71. http:// digitalcommons.unl.edu/droughtfacpub/71. WMO, World Meteorological Organization, 2016. Handbook of drought indicators and indices. Integrated Drought Management Programme, WMO-No. 1173 (52 pp). Xu, L., Samanta, A., Costa, M.H., Ganguly, S., Nemani, R.R., Myneni, R.B., 2011. Widespread decline in greenness of Amazonian vegetation due to the 2010 drought. Geophys. Res. Lett. 38:L07402. http://dx.doi.org/10.1029/2011GL046824. Yousefirad, M., 2005. The role of tectonic in the Shazand kars genesis. J. Sci. (Islamic Azad Uni.) 15 (57), 302–317. Yu, G., Yu, Q., Hu, L., Zhang, S., Fu, T., Zhou, X., He, X., Liu, Y., Wang, S., Jia, H., 2013. Ecosystem health assessment based on analysis of a land use database. Appl. Geogr. 44, 154–164. Zhao, Y.Z., Zou, X.Y., Cheng, H., Jia, H.K., Wu, Q., Wang, G.Y., Zhang, C.L., Gao, S.Y., 2006. Assessing the ecological security of the Tibetan plateau: methodology and a case study for Lhaze County. J. Environ. Manag. 80, 120–131.