Drought monitoring of south of Khuzestan province, Iran using remote sensing and SPI



Extended abstract
1- Introduction
Drought is a normal, recurrent feature of climate. It occurs almost everywhere, although its features vary from region to region (Dracup,1980). In the most general sense, drought originates from a deficiency of precipitation over an extended period of time, resulting in a water shortage for some activity, group or environmental sector (NDMC define). Consequently, it is required to demonstrate the distribution and degree of drought condition timely, which is crucial for drought warning and resisting effectively.
The impact of drought on society and agriculture is a real issue but it is not easily described. Reliable indices to detect the spatial and temporal dimensions of drought occurrences and its intensity are necessary to assess the impact and also for decision-making and crop research priorities for alleviation (Chapra,2006Seiler and Kogan et al.,1998)
Therefore, the main objectives of the study are to monitoring drought by remote sensing technique.

2-Data and Methodology:
Study Area
The case study area is south part of Khuzestan, in southwest of Iran (Fig 1). There are a variety of land covers in the area, including vegetation, bare area, and waters.

Data and method
Surface Reflectance (SR) data of the MODIS Satellite used in this study. This data have a horizontal resolution of 250m and project on a sinusoidal grid. Data have been collected during the period from 2002to 2008 to calculate NDVI. The Normalized Difference Vegetation Index is a satellite-derived global vegetation indicator obtained from the red, near-infrared (NIR) ratio of vegetation reflectance in the electromagnetic spectrum. (Weixin,2011)
NDVI = ρIR – ρR / ρIR + ρR
The parameters, ρIR and ρR, are the reflectivity of nearinfrared and red channels, respectively.
NDVI provides information on vegetation productivity and phenology over large temporal and spatial scales and has been widely used in the recent ecological studies as a proxy for vegetation productivity and phenology. NDVI is a good indicator of green biomass, leaf area index, and patterns of production. NDVI was computed using two bands of a surface reflectance image, which varies from -1 to +1.
The rainfall data during 1980-2008 were collected from 8 meteorological stations available in the province .These monthly precipitation data were checked for quality control before calculating SPI. The SPI is computed by fitting a probability density function to the frequency distribution of precipitation summed over the desired time scale. This is performed separately for each period/month and for each location in space. Each probability density function is then transformed into the standardized normal distribution (z-distribution). In this study the SPI was calculated on the 12-month time scale (for the end of December) which reflects long-term precipitation patterns. Lloyd-Hughes and Saunders (2002) describe in detail the calculation of SPI, which is outlined here. Those authors also tested the standardization procedure (probability transformation) assuming normal, log–normal, and gamma statistics for precipitation, and concluded that the gamma distribution provides the best model for describing monthly precipitation over most of Europe, especially at the 12-month scale. (costa,2011). . Gamma distribution function is:
For > 0
Where >0 is shape factor, >0 scale parameter, precipitation and Gamma function.
It can seen that the SPI is the standardize Z. SPI could be calculated for any time scale such as one month and more than it. It is generally that SPI calculated for 3, 6, 12, 24 months scale for drought monitoring.SPI values vary from -3.5 to +3.5. (table 1)

The following section describes the methodology used in the study. Correlation and regression techniques are used to verify if there is a correlation between NDVI and rainfall in south part of Khuzestan between 2002-2008. NDVI and SPI are then compared to produce the best relationship between months. A schematic presentation of the methodology that has been followed is mentioned in figure 2.
Select the appropriate place to obtain NDVI for each station
As a new approach to find the best location, climatic conditions similar to rain stations should be considered. Hence, six important factors are determined, whereas, each of these conditions will form a layer in the positioning.
The location is located about in the minimum distance from any station.
The location is in rainfall station classes.
The location should be at least in nearest distance to temperature class.
The location should be considered at least in nearest distance to elevation class.
The location should be in distance to humanity resources. (At least 1 km)
The desired location is better placed in first class areas.
So, after the layer formation the priority is based on their importance to each region and each layer is given in percent. (Table 2).
After weighting to each of the layers, six layers are combined. In fact, the following formula shows this issue as the overlap weighted in geographic information systems.. Positioning map for each area (MSS)was calculated using bellow equation(Abshirini,2009) :
MSS =(R*15) + (ds*15) + (veg*25) + (T*15) + (drh*15) + (E*15)/100
It should be noted that the considerable work is done individually for each area. Therefore, local stations in each area investigated. After that, to avoid writing more and more all selected areas be shown together on the map (Fig.3).
After positioning the best location, attempting to clip the parts selected from images is started, in which, NDVI is applied. Afterwards, means of these parts for each year is computed and the amount of this mean is used as a number in NDVI range to compare.

Rainfall data collected from Iran Meteorological organization which was used to derive Standardized Precipitation Index (SPI). Multi-scale SPI (1-month, 2-month, 3-month, 6-month, 9-month and 12-month) was calculated to detect occurrence of drought using the monthly precipitation (1980-2008) data set.
Pearson correlation analyses were conducted for the NDVI vs. 1-, 2-, …, and 12-month SPIs. Linear regression was implemented on the NDVI time series and 1 to 12-month SPI. The 12-month SPI was found to have the best correlation with the NDVI, indicating lag and cumulative effects of precipitation on vegetation, but the correlation between NDVI and SPI varies significantly between months (table 4). The highest correlations occurred during the middle of the growing season, and lower correlations were noted at the beginning and end of the growing season in most of the area.The results also indicate that the highest correlations occurred during the middle of the growing season (June), and lower correlations were noted at the beginning and end of the growing season in most of the area. Figure 2 shows the regression equation between mean NDVI of June calculated from 13 June NDVIs .there is a positive correlation between NDV and SPI, and this means that the more rainfall cause the more vegetation cover.
Finally we prepared drought map using satellite data (Fig. 5).The results shows that the Rs based map has a good presentation of drought situation of the province. Figure 5 shows the real drought map calculated based on rainfall data and SPI method in compare with remotely sensed map using NDVI.

In this study, the meteorological drought prone areas in the south-west region of Iran were identified by using Remote Sensing and GIS technology and drought risk areas were to delineate by integration of satellite images and meteorological information. The role of satellite derived index for drought detection has been exemplified by integrating meteorological derived index called Standardized Precipitation Index. It is found that the temporal variations of NDVI anomaly are closely linked with SPI and a strong linear relationship exists between NDVI and SPI. Highest correlation was found in Shushtar area with a R² value of 0.85.There is a significant correlation between NDVI anomaly and 12-month SPI. Thus,it can be said that NDVI index and precipitation index shares a strong correlation where water is a major limiting factor for plant growth. We also find that there is the highest R squares between SPI and NDVI for 6 to 12 month SPI and lower for 6, 3 and 1 month SPIs. The drought maps indicates the more frequent drought for the sought and west of the province in compare to the other parts of the studied area.