Socioeconomic and Biophysical Influences on Forest Cover type in the Alabama's eight Counties: A logistic Regression Analysis using GIS and Remote Sensing Techniques

of 15

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
PDF
15 pages
0 downs
4 views
Share
Description
Socioeconomic and Biophysical Influences on Forest Cover type in the Alabama's eight Counties: A logistic Regression Analysis using GIS and Remote Sensing Techniques
Tags
Transcript
   1 Socioeconomic and Biophysical Influences on Forest Cover type in the Alabama’s eight Counties: A logistic Regression Analysis using GIS and Remote Sensing Techniques Buddhi Gyawali, Wubishet Tadesse, Rory Fraser, Yong Wang Abstract Land use and land cover study has recently become an important subject in understanding social and economic dynamics of the landscape. The availability of census data at the finer scale and development of spatial analysis tools for analyzing raster and vector data have made such studies more practical and relevant. This study examined the relationship  between land cover and socioeconomic and biophysical characteristics of Alabama's eight counties at the census block group levels. This study analyzed U.S. census 2000 and Landsat 2000 imagery data, soil, elevation, road, and streams data using ERDAS, ArcView 3.2, and ARCGIS. The classified land cover raster image was converted to the vector format and overlaid to a vector layer of census 2000 block groups and biophysical data. This information was statistically analyzed using binary logistic regression model. The result suggested that biophysical variables are better predictors of forest cover than socioeconomic variables.  Introduction The Southwest region of Alabama consists of eight counties with the lowest Human Development Indices (HDI) in the state (Bukenya and Fraser 2003). Afro-Americans constitute over 60% of the region’s population and have the lowest share of the gross value of agriculture and timber products (USDA, 1999), the lowest employment rates in forest-based pulp and paper industries (Bailey et al. 1996), and the lowest  participation rates in non-timber forest product harvesting and recreational activities on  public lands (Johnson, 2000). Social theory based studies suggest a correlation between natural resource and Afro-Americans shaped by resource dependency (Schelhas and Zabawa 2000; Bliss et al. 1998), power relations and land tenure (Schulman 1991), labor structure (Bailey et al. 1996), race and class based discrimination (Mitchell 2001). All of these studies have recognized that upheavals in black landownership and tenancy have disconnected Afro-American involvement in the natural resources section (Gilbert et al., 2001; Gavanta, 1995; Ayers, 1992). However, none of these studies has explicitly addressed the spatial dimensions of the “connection between people, poverty, and natural resources (Wallace and Knight 1996), which is considered an important ‘subject of study’ in the contemporary research (Fox et al., 2003, Moran et al., 2003). Studies in global  poverty and economic development have analyzed this ‘connection’ and found spatially explicit explanations for poverty (Leslie, 2000). Such research has shown the need for ‘level of analysis at fine scale’ in order to better understand of the factors that shape the relationship between people and the available economic and natural resources (Rindfuss et al., 2003). The proposed study therefore, utilizes spatial data at a finer geographic scale to analyze the relationship between people and the resources upon which they depend on for their livelihood (Goodchild, 2004). Since the economy of the black belt region is  primarily based on natural resource activities (agriculture and natural and plantation   2forests), the relationship between available resources and the people who are entitled to use these resources is important for understanding the poverty of the study area. Non-spatial studies of these counties have indicated that landscape characteristics (type of landownership and land cover, location of the property, population distribution, and road network) play an important role in how people have benefited from agricultural and forest resources (Joshi, et al., 2000, Bliss et al.,1998). A Recent study, which was conducted using the information available at the census block groups, found that the type of land cover of the black belt counties is correlated with population density, poverty, and Afro-American populations (Gyawali et al., 2004). However, studies in different parts of the world suggest that other non-socioeconomic variables such as type of soil, elevation, slope, water availability and channels, and road networks are significantly correlated with the agricultural and forest resources (Fox et al., 2003; Moran et al., 2003). In this research, socioeconomics and biophysical variables are examined at the census block group (CBG). The CBGs are chosen ‘as the unit of analysis’ because these are the lowest unit for which the most important socioeconomic information are readily found. Theoretical Framework, Selection of Methodology and Variables This research is guided with the perception that poor and socially marginalized  people live in the geographic and biophysical conditions which are considerably different than other populations. In the literatures of ‘spatial study of the global poverty research’, the constraints such as access (to roads, railroads, and coast), climate (precipitation, growing season, climate zones, droughts), demography (density and urbanization), topography (elevation and slope), soil quality, disease (e.g. malaria), as well as water availability conditions are identified and assessed. Such data is aggregated into the standard grid and analyzed using multiple regression or discrete regression models (such as logit or probit) to identify which variables explain poverty, what combination of the variables generates the best fit, and what proportion of the variance is accounted by social characteristic and biophysical characteristics. Landscape ecology related studies have found the relationships between socioeconomic change and land use changes (which is considered a major indicator of economic development) (Moran et al., 2003). The land use change is induced through favorable ecological characteristics (richness of high economic value land and forest species, water resources, and fertile soil, favorable road network and slope), which attracts more investors and increases human population with greater opportunities for natural resource -based activities. For instance varying elevation and erratic distribution of topography, climate, and soil may cause variation of forest species thereby causing varying net primary productivity of forest products. Studies also suggest that poor transportation access, poor agricultural conditions (sandy soil, drought), and inefficient geography (territories smaller, farther apart residents) have a significant relationship in causing high poverty. For instance, Axinn and Barber (2003) found that size of investment in infrastructure plays an important role in poverty alleviation as it opens the opportunity for investment. Roads create accessibility to the resources and assists in transportation of logs, agricultural goods and farm inputs. Roads  provide an opportunity for the farmers to involve in a wider range of economic activities (Gibson and Rojelle, 2003). Access to roads affects the price farmers receive from selling the crops and price they pay for purchased consumer goods. Distance between the point   3at which people can get access to roads and the center of the government offices (center of economic activities) is also relevant. Other studies (Walsh et al., 2003) found that increasing distance of roads decreases diversification of economic activity (number of income earning activities). The study by Gibson and Rojelle (2004) shows that poverty is dominant in rural area, and is related to those communities that have poor access to services, markets, and transportation. Studies have found that economic status of households living in steep slope areas are comparatively poorer than the households living in the flat areas. The lands in the steep slope are subject to inundation and suffer periodic shortfall of the rainfall. These factors may create difficulty in building roads (Pascual and Edward, 2004). Steep slope, waterlogged and inundated soils can also reduce the agricultural potential (low acreage of croplands). Soil degradation due to poorly drained or unsuitable soils for particular crops affects agricultural productivity, which is manifested through their impacts on variances in yield and total productivity of agricultural goods. These impacts increase economic costs in the form of loss of income, increased risk, and increased costs of production. Moran et al. (2003) suggest that natural forests are associated with high infiltration rate and low soil erosion indicating lower numbers of streams. However  plantation/secondary forests do not necessarily have these characteristics due to the higher number of roads, logging activities, artificial drainage ditches, etc. Fox et al. (2003) discovered that forest management activities may increase flood due to cultivation, road construction, increased stream density, and soil compaction during logging. Profile of the Study Area The major socioeconomic and biophysical characteristics of the study region, which are summarized at CBG level are presented in the Table 1. The study site (-86.4 to-88.4 degree E, 31.13 to 33 degree North) consists of eight counties (Dallas, Green, Hale, Lowndes, Marengo, Perry, Sumter, and Wilcox) located in the southwest part of Alabama. The area covers 6,479 square miles (4,197,125 acres). The region is called black-belt  because of the predominant African American population and presence of the black calcareous soil. The total population of the region is 149,378 of which 65% are Afro-Americans (U.S. Census 2000). The population density is 22 people per square mile. Thirty two percent of the people live with income below the national poverty standard. The record of medium household income and bachelor graduates for whites is two times greater than the same records for the Afro-Americans (U.S. Census 2000). The mean elevation of CBG in the eight counties is 180 meter above the sea level. The road and stream channel density is found to be 3.67 and 5.13 meter per acre in the CBGs. The average percentage of forest and crop/pasture cover in the census block groups is 48.41% and 29% respectively (Gyawali et al., 2004). The major forest tree species are loblolly, oak-pine, oak-hickory, longleaf slash pine, and oak-gum cypress. The major agricultural crops are peanuts, soybean, and hay crops. The study area is mostly rural and population mostly depends on forest-based industry, agricultural, and livestock employment. The lower percentage of people employed (47.78 for whites and 37.92 for blacks) indicates that most of the people do not have a full-time job.   4 Table 1. Descriptive Statistics of Socio-economic and Bio-physical variables Characteristics Mean (N = 161) Median Std. Deviation Standard Error Total acres 24419.8 14749.44 28931.91 2280.5 Crop% 29.03 25.95 24.78 1.95 Forest% 48.41 54.23 30.27 2.38 Elevation 179.57 179.6 .94 .0075 Road (m/acre) 3.67 2.187 3.91 .30 Creek/acre (m/acre) 5.13 5.00 2.32 .18 Total Population 928 841 406.27 32.02 Population density (people/acre) 1 1 1.83 .14 Whites (%) 33.58 26.68 26.30 2.07 African Americans (AA) (%) 65.40 72.93 26.83 2.11 Population below poverty(%) 31.64 31.87 14.14 1.11 White High school Graduates (%) 32.63 33.79 17.48 1.37 White Bachelor Graduates (%) 10.90 9.59 9.440 .74 AA High School graduates (%) 33.38 32.86 13.43 1.05 AA Bachelor Degree (%) 4.27 3.16 4.26 .33 White Employed (%) 47.78 51.87 19.53 1.53 AA Employed (%) 37.92 36.44 13.02 1.02 White Medium Household Income ($$) 36276.10 36793 18366.69 1447.50 Black Medium Household Income ($$) 17543.09 15528 9007.20 709.87 White Per capita Income ($$) 21136.66 19660 14591.99 1150.01 AA Per capita Income ($$) 9003.50 8127 3361.26 264.90 Data Sources Four sets of data were used in the study. These were: Census 2000 block group vector layer, Landsat 2000 ETM+  image, Census 2000 socioeconomic data, and biophysical data. A. U.S. Census 2000 Block group shape file Eight African-American dominant counties were selected from the southwestern region of Alabama (Figure 1). Census block groups layers of each county were downloaded from census 2000 data base and merged together utilizing ArcView’s Geoprocessing extension. There were 161 block groups (polygons) with varying shape. The average area of the CBGs is 24,419 acres. These CBGs were considered as ‘the unit of analysis’ for extraction of other socioeconomic and biophysical data.   5 B. Landsat 2000 ETM Data Landsat Enhanced Thematic Mapper (ETM+) satellite image of 2000 was used to derive different land use types. The Landsat data recorded in September 2000, and was rectified, terrain corrected, and geo-referenced to local UTM zone (WGS84 Datum). The  positional accuracy was ± 50 meters RMS. The black-belt region required three scenes (Path/row: 20/38, 21/37, and 21/38) which were combined to create a mosaic of the study area. A vector layer of the UTM projected boundary map of eight black-belt counties was used to create a subset of landsat image for image classification. The Landsat ETM+  image was preprocessed using ERDAS IMAGING 8.6 software to increase the contrast and brightness using a principle component analysis. The total number of six bands was reduced into three components. The new image had a  better view for image classification. The 2000 image was initially used to extract 15 classes based on clustering algorithm in unsupervised classification using the Anderson level 2 classification schemes. Since the resolution of this data was 28.5 meters, it wasn’t possible to conduct higher level classification (level III or IV) (Jensen, 1996). For this reason, the residential Figure 1 (b) A census block group layer of the study area Figure 2 (a) Landsat ETM +  subset image of the study area Figure 2 (b) An overlay of classified image and census block group layer Figure 1 (a) Study area showing Eight counties in southwest region of Alabama
Related Search
Similar documents
View more...
Advertisements
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks