Influence of Vegetation Changes on Urban Flooding, Case Study, Behshahr Flood Catastrophe
Abdolmotalleb Rastegar; Amin Maldar Badeli; Mohammad Ali Nezammahalleh*; Seyed Kazem Alavipanah; Abdolhossein Hajizadeh
Department of Engineering and Technology, Golestan University, Aliabad Katool, Iran
Master of Civil Engineering, Department of Civil Engineering, Neka Branch, Islamic Azad University, Neka, Iran
* PhD Student of Geomorphology, Physical Geography Department, University of Tehran, Tehran, Iran
()
Prof. in Remote Sensing and GIS, Department of Remote Sensing and GIS, University of Tehran, Tehran, Iran
Abstract
Urban flooding has obviously considerable consequences for human societies. According to evidence both urbanization and intensity of rainfall due to climate change is rapidly increasing. This may lead to a rise in probability of urban flooding. The purpose of the study is to explore vegetation changes in Behshahr Watershed by processing Remote Sensing data to understand whether the vegetation changes could be effective in intensification of the catastrophic flood event of October, 2012, in Behshahr City. Landsat data in 1987, 2001, and 2006 have been processed to give NDVI maps. The NDVI maps were subtracted from each other to detect changes. The results indicated that a decrease in vegetation density from 1987 to 2006 has exacerbated the effects of intense rainfall and lead to the hazardous event in Behshahr City.
Keywords: urban flooding, vegetation changes, NDVI, Behshahr
Introduction
As the population and use of natural resources is increasing as well as development of new technologies make it possible for human being to destroy the environment intentionally or unknowingly, the challenges for management of the environment is even more complicated (Gregersen, 2007, p. 1). Many natural hazards are a normal behavior of nature that may be exacerbated by human interventions in natural system. Flooding is one of the natural behaviors that may be considered as a hazard because of the human intervention or the lack of knowledge of society about their characteristics. Changes in vegetation and habitats (Ghahroudi, 2012) in watersheds aresome of the interferences that initiate and intensify the floods (Gregersen, 2007, p. 23).Urban settlements are influenced by these floods with lots of damages throughout the world (Chen, 2009; Jochen, 2012; Ghahroudi and Nezammahalleh, 2013). Urban flooding is mainly resulted from decrease in vegetation cover, development of impermeable surfaces in watersheds, population growth and advance of urban structures into flood prone areas, climate change with increase in precipitation intensity, and the old building blocks in urban areas (Jochen, 2012). In one hand, the urbanization is highly increasing so that 49.2 percent of people now are living in urban areas (Chen, 2009) and the number will be about 70 percent by 2050 (Chen, 2012). On the other hand, the climate change promotesthe intensity of precipitations as the main cause of the urban flooding (Beygloo, et al., 2013). A decrease in vegetation can lead to higher runoff values and also an increase in the speed of runoff and discharge of rivers. Vegetation cover in watershed areas mainly intercept the rainfall effect on the soil and allow slow infiltration of the runoff and prevent a huge flow of runoff in a limited time interval. Therefore, the higher vegetation covers can reduce the impacts of precipitation and consequently the flooding.
According to the recent statistics up to 72 floods occurred in Iran from 1900 to 2012 that left 7767 people dead and 3652534 people affected and had 7652528 $ financial losses for the country. From all the annual budget of Unexpected Events Staff 70% was allocated for the losses of flooding. According to the statistics from the staff during the past 50 years Iran has experienced 2400 horrible floods that inundated 480000 buildings with many human and financial casualties (Ghahroudi and Nezammahalleh, 2013). The urban flooding in 14th October, 2012, in Behshahr, Iran, has caused lots of damages. Up to 235 buildings, 90 automobile vehicles, 50% of agricultural lands, and roads were nearly destroyed and 8 people died by the catastrophe (Begloo, et al., 2013).The purpose of the study is to investigate the effects of vegetation changes in upstream watersheds of the City.
The study area of the research is about 210 Km2 in the watersheds of the streams that discharge into the city (Figure 1). The Bosghool River as the main drainage into the city, with an average discharge of 6 M3/s, experienced a discharge of 90 M3/s at the date of the event. A rainfall with 90 mm in 3 hours was reported.
Figure 1: position of the study area
Materials and Method
There are many indices to detect changes in vegetation and also vegetation areas (Alavipanah, 2003). One of the most important is Normalized Difference Vegetation Index (NDVI). This is the difference of NIR and Red bands divided by the sum of them. The index can reveal changes in vegetation over time and also classify vegetation covers based on their concentration in an area.
NDVI map have been prepared by Landsat Images of TM1987, ETM+2001, and ETM+2006. After preprocessing and needed corrections, the images were prepared for analyses. Following normalization in ArcGIS, the NDVI map of 2001 was subtracted from NDVI map of 1987 and the map of 2006 from 2001 to give the differences in vegetation classes. The difference between three raster layers, as the recent image has been subtracted from the previous, has indicated the vegetation changes during the two periods.
NDVI1987-NDVI2001=DIF1
NDVI2001-NDVI2006 =DIF2
We hypothesized that a decrease in vegetation caused the flooding. In NDVI maps as the pixel values proximate to 0 it represents a decrease in vegetation and as it proximate to 1 it represents an increase in vegetation density. Thus, NDVI of the newer image must be more intense in vegetation in most of the pixels. In other words, the value of a pixel in older image must be more than that of the newer image. If the pixel value of newer image is more than that of older, the value of that pixel in the result map (DIF1, DIF2) will be a minus quantity. Therefore, in the maps of DIF1 and DIF2, the areas with minus pixel values are indicative of increase in vegetation density.For an easy interpretation, the results have been classified in three classes of increase in vegetation, no considerable changes and decrease in vegetation.
Results and discussion
As mentioned the maps have been classified in three categories. The classification has been done based on natural breaks method in ArcGIS Spatial Analysis. The results have indicated that vegetation has been decreased from 1987 to 2001 for an area up to 2.87 km2, mainly in upstream areas. Up to 186 km2 had no considerable changes. About 20 km2 have experienced an increase in vegetation, mainly in low land areas.In DIF1 the area of decrease is negligible but it is mainly in high watershed areas where it directs runoffs into tributaries (Figure 2). As it can be seen in the figure mainly upstream areas have experienced some decline in vegetation density and the areas in the moderate portions of the study area have not undergone any considerable changes during the period of observation. In the northern parts of the area that is greatly under the influence of settlement regions with many rural areas some increase can be observed in vegetation concentration. The augment in vegetation cover may be the growth of trees in urban areas and some changes in agricultural land uses.
Figure 2: changes of vegetation from 1987 to 2001
In DIF2 from 2001 to 2006 about 1 km2 have increased in density of vegetation. Nearly 202 km2of the area have no considerable change in vegetation and about 4.5 km2 have decrease in vegetation. The pixels with decrease are dispersed all over the watershed and they can be seen mainly along river flood plains(Figure 3). As it is observable from the figure, there is not considerable decrease in concentration of vegetation cover in the study area. From 2001 to 2006 the areas that have experienced some decreases in vegetation are generally dispersed over the study area. The decrease in vegetation is not, similar to the previous period, concentrated to large extent in highland mountain areas or in settlement areas, but the pixels with decrease in vegetation can be recognized in all the study area as dispersed. This may be due to changes in climate or a natural large scale factor that influenced all the surrounding areas with the same level and to a certain magnitude. As it can be seen the decreases in vegetation have mainly occurred along the fluvial channels and the tributaries. This can intensify the influence of precipitation intensity and also the effects of discharge flows.
Figure 3: changes of vegetation from 2001 to 2006
Conclusion
Today, one of the most optimized and streamlined method to mitigate the devastating consequences of natural hazards is an adaptation with the environment in such a way that living of human being on the earth has the least adverse impacts on her nature. For this, it is required to know behavior of the nature and the internal relationship among its components for a prediction of the behavior. This can be concluded from the study that along with the increase in urbanization and in the intensity of precipitation and consequently higher probability of urban flooding, the vegetation as a factor preventing flooding is decreasing progressively. The high precipitation in short time interval has played the main role in this event and the decrease in vegetation was greatly as exacerbation.
References
(1). Alavipanah, S.K. (2003) Application of remote sensing in the earth sciences (soil). University of Tehran Press, 173
(2). Chen S. A.; Evans B.; Djordjevic S.; Savic, A.D., (2012) Multi-layered coarse grid modeling in 2D urban flood simulations. Journal of Hydrology 470-471, 1-11
(3). Ghahroudi Tali, M., Nezammahalleh, M.A., (2012) Urban flooding management using the natural drainage system case study: Tehran, capital of Iran. IAHS-AISH publication, 174-180
(4). Ghahroudi Tali, M., Sadough, S.H., Nezammahalleh, M.A., Nezammahalleh, S.K., (2012) Multi-criteria evaluation to select sites for ecotourism facilities: a case study Miankaleh Peninsula. Anatolia 23, 373-394
(5). Gregersen, Hans, M.; Ffolliott, Petter, F.; Brooks, Kenneth, N.; (2007) Integrated watershed management, connecting people to their land and water. CABI international, 1, pp 1-23
(6). Jafar Beygloo, M.; Nezammahalleh, M.A.; Hajizadeh, A.H.; Bayati Sedaghat, Z., (2013) Effects of watershed Physiographic characteristics on intensification of urban flooding in Behshahr. 1st National Conference of Geomorphology (in Farsi)
(7). Jian Chen, Arleen A. Hill., Lensyl D. Urbano, (2009) A GIS-based model for urban flood inundation, journal of hydrology 373, 184-192
(8). Jochen E. Schubert; Brett, F.S., (2012) Building treatments for urban flood inundation models and implications for predictive skill and modeling efficiency, advances in water resources 41, 49-64