Monitoring urban growth and detecting land-cover changes on the Istanbul metropolitan area

Istanbul is the most populated city of Turkey with a population of around 10.58 M (2000) living on around 5,750 km². In 1980, the population was only 4.7 M and then it has been more than doubled in only two decades. The population has been increasing as a result of mass immigration. An urbanization process continues and it causes serious increases in urban areas while decreasing the amount of green areas. This rapid, uncontrolled, and illegal urbanization accompanied by insufficient infrastructure has caused degradation of forest and barren lands in the metropolitan area, especially through the last two decades. The watershed basins inside the metropolitan area and the transportation network have accelerated the land-cover changes, which have negative impacts on water quality of the basins. Monitoring urban growth and land cover change will enable better management of this complex urban area by the Greater Istanbul Metropolitan Municipality (GIMM). A temporal assessment of land-cover changes of Istanbul has been documented in this study. The study mainly focuses on the acquisition and analysis of Landsat TM and Landsat GeoCover LC satellite images reflecting the significant land-cover changes between the years of 1990 and 2005. Raster data were converted to vector data and used in Geographic Information Systems (GIS). A database was created for Istanbul metropolitan area to plan, manage, and utilize statistical attribute data covering population, water, forest, industry, and topographic position. Consequently an overlay analysis was carried out and land use/cover changes through years have been detected for the case study area. The capability of Landsat images in determining the alterations in the macro form of the city are also discussed.     

Emergence of Indigenous Vegetation Classifications Through Integration of Traditional Ecological Knowledge and Remote Sensing Analyses

Traditional ecological knowledge (TEK) can play an important role in the understanding of ecological systems. Although TEK has complemented scientific and managerial programs in a variety of contexts, its formal incorporation into remote sensing exercises has to date been limited. Here, we show that the vegetation classifications of the Ache, an indigenous hunter-gatherer tribe of the Mbaracayu Forest Reserve in Paraguay, are reflected in a supervised classification of satellite imagery of the reserve. Accuracy of classification was toward the low end of the range of published values, but was reasonable given the difficult nature of separating forest classes from satellite images. Comparison of the resultant map with a more traditionally elaborated vegetation map highlights differences between the two approaches and the gain in information obtained by considering TEK classifications. We suggest that integration of TEK and remote sensing may provide alternative insights into the ecology of vegetation communities and land cover, particularly in remote and densely forested areas where ecological field research is often limited by roads and/or trail systems.     

Landsat7 ETM+ SLC-OFF数据的修复及其在武汉东湖水质反演中的应用

自从2003年5月31日,从陆地卫星 7发回的ETM+图像数据就存在缺陷。这是由于增强专题制图仪的扫描线校正器发生故障引起的。这些称为SLC OF数据的图像有一些黑色的不存在任何数据的扫描行。丢失的数据约占全景数据的25％,使它们难以正常使用。但是,数据本身仍然保留了良好的辐射和几何性质,如加以妥善修复,仍可以在一些特殊领域中使用。首先介绍了如何使用自适应局部回归算法(ALR)恢复这些图像,然后使用修复后的图像反演武汉东湖的水质参数。结果表明：ALR算法可以对SLC OF图像进行较好的修复,而且利用修复后的图像和东湖的地面水质监测数据,通过多元逐步回归分析,可以建立很好的叶绿素a、透明度、总磷以及总氮等水质参数的经验遥感反演模型,模型的相关系数R²分别为0.86、0.75、0.73和0.71。反演得到的水质参数分布与实际情况符合。这些数据有许多优点,如空间分辨率高、存档数据非常丰富、可以从NASA的服务器免费下载等。在其他遥感数据不足或无法获得的情况下,这些数据经过适当的修复,可以作为补充或替代数据使用。〖     

Using Spatial Pattern to Quantify Relationship Between Samples, Surroundings, and Populations

The need for accurate carbon budgeting, climate change modelling, and sustainable resource management has lead to an increase in the number of large area forest monitoring programs. Large area forest monitoring programs often utilize field and remotely sensed data sources. Sampling, via field or photo plots, enables the collection of data with the desired level of categorical detail in a timely and efficient manner. When sampling, the aim is to collect representative detailed data enabling the statistical reporting upon the characteristics of larger areas. As a consequence, approaches for investigating how well sample data represent larger areas (i.e., the sample neighbourhood and the population) are desired. Presented in this communication is a quantitative approach for assessing the nature of sampled areas in relation to surrounding areas and the overall population of interest. Classified Landsat data is converted to forest/non-forest categories to provide a consistent and uniform data set over a 130,000 km² study region in central British Columbia, Canada. From this larger study area 322 2 × 2 km photo plots on a 20 × 20 km systematic grid are populated with composition and configuration information for comparison to non-sampled areas. Results indicate that typically, within the study area, the spatial pattern of forest within a photo plot is representative of the forest patterns found within primary and secondary neighbourhoods and over the entire population of the study. These methods have implications for understanding the nature of data used in monitoring programs worldwide. The ability to audit photo and field plot information promotes an increased understanding of the results developed from sampling and provides tools identifying locations of possible bias.     

The Impact of Landsat Satellite Monitoring on Conservation Biology

Landsat 7s recent malfunctioning will result in significant gaps in long-term satellite monitoring of Earth, affecting not only the research of the Earth science community but also conservation users of these data. To determine whether or how important Landsat monitoring is for conservation and natural resource management, we reviewed the Landsat programs history with special emphasis on the development of user groups. We also conducted a bibliographic search to determine the extent to which conservation research has been based on Landsat data. Conservation biologists were not an early user group of Landsat data because a) biologists lacked technical capacity – computers and software – to analyze these data; b) Landsats 1980s commercialization rendered images too costly for biologists budgets; and c) the broad-scale disciplines of conservation biology and landscape ecology did not develop until the mid-to-late 1980s. All these conditions had changed by the 1990s and Landsat imagery became an important tool for conservation biology. Satellite monitoring and Landsat continuity are mandated by the Land Remote Sensing Act of 1992. This legislation leaves open commercial options. However, past experiments with commercial operations were neither viable nor economical, and severely reduced the quality of monitoring, archiving and data access for academia and the public. Future satellite monitoring programs are essential for conservation and natural resource management, must provide continuity with Landsat, and should be government operated.     

RIDGE REGRESSION TECHNIQUES APPLIED TO LANDSAT INVESTIGATION OF WATER QUALITY IN LAKE OKEECHOBEE1

ABSTRACT: Landsat radiance values were processed at two different (single and double) levels of accuracy to estimate chlorophyll a, turbidity, and suspended sediment in Lake Okeechobee, Florida. Both ordinary least square and ridge regression analyses were used to establish a relationship between water quality parameters and Landsat radiance. Radiance measurements made at greater precision (double level) gave a better solution in this application. The ridge regression analysis for double level not only can reduce the total mean square error about 13–20 percent and confidence interval about 6–28 percent as compared to ordinary least square analysis, but it can also change the interpretation of analysis results.     

Mapping Disturbances in a Mangrove Forest Using Multi-Date Landsat TM Imagery

To evaluate the accounts of local fishermen, Landsat TM images (1986, 1993, 1999) were examined to assess potential losses in the mangrove forests of the Teacapán–Agua Brava lagoon system, Mexico. A binary change mask derived from image differencing of a band 4/3 ratio was employed to calculate any changes within this forested wetland. The results indicate that by 1986 approximately 18% (or 86 km²) of the mangrove area under study was either dead or in poor condition. The majority of this damage had occurred in the eastern section of the Agua Brava basin, which coincides, with the reports of the elderly fishermen. Examination of aerial photographs from 1970 revealed no adverse impacts in this area and would suggest, as postulated by the fishermen and other scientists, that modifications in environmental conditions following the opening of a canal, Cuautlá canal, in 1972 may have initiated the large-scale mortality. Although these areas of impact are still developing, the results from the satellite data indicate that the majority of the more recent changes are occurring elsewhere in the system. Obvious in the 1999 satellite data, but not so in the 1993, are large areas of mangrove degradation in the northern section of the Teacapán region. In the Agua Brava basin, the more recent transformations are appearing on the western side of the basin. Since long-term records of environmental conditions are absent, it is difficult to determine why these latest changes are occurring or even if the earlier losses were the result of the canal. Potential agents of change that have recently been observed include a hurricane, a second canal, and the uncontrolled expansion of the Cuautlá canal since 1994.     

Sampling and Mapping a Soil Erosion Cover Factor by Integrating Stratification,Model Updating and Cokriging with Images

Cost-efficient sample designs for collection of ground data and accurate mapping of variables are required to monitor natural resources and environmental and ecological systems. In this study, a sample design and mapping method was developed by integrating stratification, model updating, and cokriging with Landsat Thematic Mapper (TM) imagery. This method is based on the spatial autocorrelation of variables and the spatial cross-correlation among them. It can lead to sample designs with variable grid spacing, where sampling distances between plots vary depending on spatial variability of the variables from location to location. This has potential cost-efficiencies in terms of sample design and mapping. This method is also applicable for mapping in the case in which no ground data can be collected in some parts of a study area because of the high cost. The method was validated in a case study in which a ground and vegetation cover factor was sampled and mapped for monitoring soil erosion. The results showed that when the sample obtained with three strata using the developed method was used for sampling and mapping the cover factor, the sampling cost was greatly decreased, although the error of the map was slightly increased compared to that without stratification; that is, the sample cost-efficiency quantified by the product of cost and error was greatly increased. The increase of cost-efficiency was more obvious when the cover factor values of the plots within the no-significant-change stratum were updated by a model developed using the previous observations instead of remeasuring them in the field.     

Land-cover change in upper Barataria Basin estuary,Louisiana, 1972-1992: increases in Wetland area

The Barataria Basin, Louisiana, USA, is an extensive wetland and coastal estuary system of great economic and intrinsic value. Although high rates of wetland loss along the coastal margin of the Barataria Basin have been well documented, little information exists on whether freshwater wetlands in the upper basin have changed. Our objectives were to quantify land-cover change in the upper basin over 20 years from 1972–1992 and to determine land-cover transition rates among land-cover types. Using 80-m resolution Landsat MSS data from the North American Landscape Characterization (NALC) data archive, we classified images from three time steps (1972, 1985, 1992) into six land-cover types: agriculture, urban, bottomland hardwood forest, swamp forest, freshwater marsh, and open water. Significant changes in land cover occurred within the upper Barataria Basin over the study period. Urban land increased from 8% to 17% of the total upper basin area, primarily due to conversions from agricultural land, and to a lesser degree, bottomland forest. Swamp forest increased from 30% to 41%, associated with conversions from bottomland hardwood forest and freshwater marsh. Overall, bottomland forest decreased 38% and total wetland area increased 21%. Within the upper Barataria, increases in total wetland area may be due to land subsidence. Based on our results, if present trends in the reduction of bottomland forest land cover were to continue, the upper Barataria Basin may have no bottomland hardwood forests left by the year 2025, as it is subjected to multiple stressors both in the higher elevations (from urbanization) and lower elevations (most likely from land subsidence). These results suggest that changes in the upper freshwater portions of coastal estuaries can be large and quite different from patterns observed in the more saline coastal margins.     

The use of Landsat multispectral scanner data for the analysis and management of flooding on the river Severn,England

Remote sensing has emerged as one of the major techniques for the analysis and delineation of large floods. This analysis can provide data invaluable for the hydrological management of large river systems. A need for information on the extent of floodplain inundation for the lower reaches of the largest river in the UK was met by a search through Landsat images of floods and the analysis of the best example recorded. Automated classification of the Landsat imagery of this flood on the river Severn in 1977 was used to provide estimates of the extent and spatial distribution of inundation. Flood images were generated using the Plessey IDP 3000 image processor, and the maps derived accorded well with aerial photography and qualitative flood information. Three distinct floodplain environments were delineated and flood images produced by different spectral bands compared. Specific questions prompted by flood hazard management and concerning the processes and extent of flooding were answered by the Landsat data analysis. Management of the flood risk of large rivers is expensive and remote sensing data is a relatively cheap and effective way of monitoring control works and providing data for the prediction of the effects of future hydrological works. Remote sensing is a practical way in which spatial information concerning the behavior of large dynamic systems can be obtained both quickly and relatively cheaply.