Mapping of Drought Impacts on Vegetation Cover in Yazd-Ardakan Plain Using Remote Sensing

Introduction
Generally, yearly variability of precipitation make rangelands unsuitable for crop production, and livestock grazing presents a sustainable means of food and fiber production. Droughts which often occur within a long period, may drastically affect plant community composition and may make rangelands more susceptible to diseases, insect pests, weed invasions, and overgrazing. Therefore, continuous monitoring of rangeland status would help decision makers to do necessary interventions and proper reactions against drought spells. In this regard, current ground-based methods seem to be not suitable for rangeland monitoring and regional assessment of drought impacts on its vegetation cover and seasonality. These methods always consume a plenty of time and expense. Conversely, satellite remote sensing (RS) techniques can provide useful information about changes of rangeland status due to changes in climatic attributes of the region. This technique has significant promise for development of more reliable and economically feasible measures of vegetation status over large areas. This technique can be used to monitor vegetation cover status both in time and space. In fact, monitoring of large areas at low cost and time is one of the benefits of remote sensing technique.
Methodology
The current study intends to assess the impact of climatic conditions on rangeland vegetation cover in Yazd-Ardakan plain in Central Iran. Total area of Yazd-Ardakan plain is 11630 square kilometers with elevations between 3240 and 990 m above sea level. For the assessment, an indicator namely Index of Drought Impact (IDI) was introduced. This indicator uses a remotely sensed Vegetation Index (VI) as a proxy for vegetation cover and relates the selected VI to a suite of climatic data to define relationships. In this study 36 cloud free images were taken for derivation of ten well-known vegetation indices. These images were acquired from the MODIS sensor of Terra satellite in first decade of new century (2000 to 2009). Additionally, a Landsat ETM+ image was used for preparation of land use map and also interpretations due to its good spatial resolution (30 m).
Identifying relationships between climatic attributes and rangeland vegetation status was then accomplished. In this stage, records of eight meteorological stations were collected for 2000-2009 time period. Precipitation (P), temperature (T), relative humidity (H), wind speed (W) and standardized precipitation index (SPI) were selected as the main climatic indicators of the region. Correlations between yearly values of these climatic parameters and different vegetation indices were established using R2 correlation matrix. Through this procedure, fair correlations were investigated and therefore, the best vegetation index was identified. This index was then used for monitoring of drought impact on Yazd-Ardakan rangeland vegetation. Digital Elevation Model (DEM) was also employed as ancillary data for generation of temperature and precipitation maps.
In this study, SAVI was found as the most preferred MODIS-derived vegetation index and therefore combined with precipitation and temperature, as two main climatic parameters of the region to generate the IDI map. Based on the rationale relationship in the IDI, regions with high values of IDI would be more susceptible for degradation of their resources (vegetation cover) as direct impacts of main climatic attributes in the region. In this case study, IDI map has been generated and classified into three different levels (low, medium and high) and then interpreted, visually.

Results and Discussion
Results show that in the northern areas of the studied region which are also the plain outlet have the maximum IDI values compared with other areas. These areas can be found as more susceptible regions to degradation of their resources (vegetation cover) due to climate inconsistencies and droughts in the recent decade. Lower part of Yazd-Ardakan plain (outlet) is a region with minimum agricultural and pastoral activities due to its degraded soil and water resources, salinization, harsh climatic conditions and poor natural vegetations. In contrast, lower values for “IDI” belong to border areas. These areas are high elevated regions with relatively cold winters and moderate summers. This indicator was also classified into three degradation classes to show the spatial extent of climatic impacts on rangeland vegetation cover. Results show that about 23%, 59% and 18% of this area belongs to low, medium and high IDI values, respectively.

Conclusion
Results indicate that the IDI provides an approach to connect the climate conditions with vegetation cover to assess vegetation degradation in such arid environment. This index can be used by researchers; decision makers and stakeholders for doing necessary interventions and for the purpose of sustainable rangeland management. It can also be interpreted and used based on other related aspects such as grazing pattern, distance to water and population centers. Considering results of this case study, the introduced indicator was found suitable for monitoring of drought impact on rangeland vegetation and its classification. Repeatability and relatively low cost and time consumption are the major advantages of the introduced indicator.