Multispectral Scanner Data Research Articles

Multitemporal and Multispectral Remote Sensing Approach for Flood Detection in the Elbe‐Mulde Region 2002

AbstractThe extreme flood event of the Elbe and Mulde Rivers in August 2002 caused, with its very sharp rise and decline of the water masses, a manifold of damage to the infrastructure as well as toxic deposits in the flooded areas. The paper gives a summary of the processing of multitemporal and multisensor remote sensing data for the purpose of monitoring this event. Methods and results for several thematic questions are presented and discussed. Procedures for the detection of flooded areas with operational satellite data (Landsat‐7 ETM, IRS) are investigated. Further studies are performed regarding the classification of moisture classes in the flood plain in relation to the high water changes, the accumulation of sediments and silts for different land use classes and the erosive impact of the flood by using data with high spatial resolution from satellite (IKONOS) and from aircraft (Daedalus‐Scanner).The time series of multispectral scanner data are very well suited for the spatial analysis of the damage and ecological risk, caused by the flood. The results correlate well with the data from in‐situ measurements. At the same time they show how the gathering of samples can be optimized by using remote sensing data. The results can be directly integrated in existing GIS of the public authorities for the assessment and management of extreme flood events.

Mapping of water constituents in Lake Constance using multispectral airborne scanner data and a physically based processing scheme

A physically based method to determine water constituents in case II waters from imaging spectrometer data is presented. The processing scheme includes modules for sunglitter correction, aerosol retrieval, atmospheric correction, and the retrieval of water constituents from subsurface reflectance spectra in case II waters. For the aerosol retrieval, the sun glitter, and the atmospheric correction modules, an azimuthally resolved radiative transfer model for a multilayer atmosphere-ocean system with a flat water surface is used. The retrieval of water constituents is achieved by an iterative fitting algorithm to adjust modeled and measured underwater reflectances. The processing scheme is applied to airborne multispectral scanner Daedalus AADS1268 in ATM (airborne thematic mapper) modus. Owing to the sensor properties (reduced radiometric sensitivity of the blue channel), it was not possible to work completely without additional data; Gelbstoff absorption could not be retrieved. It had to be inputted and was kept constant over the mission area. The processing system has been tested for multitemporal mapping of water constituents at Lake Constance. The results show the system as an efficient and reliable tool for the inversion of remote sensing data from natural waters.

Monitoring vegetation changes in Pasvik (Norway) and Pechenga in Kola Peninsula (Russia) using multitemporal Landsat MSS/TM data

During the period 1973–1999, the effects of air pollution on the terrestrial ecosystems in the border areas of Norway and Russia have been investigated by the use of satellite remote sensing. Landsat Multispectral Scanner (MSS) and Thematic Mapper (TM) data, in combination with field work data collected in the period 1970–2000, were applied to produce land cover maps for seven different years, with an overall accuracy of 75–83%. On the basis of this monitoring, we can conclude that the main effect of air pollution was that the areas of lichen ( Cladina spp.)-dominated forests and mountain heaths were reduced from 37% in 1973 to 10% in 1994, followed by a slight increase to 12% in 1999. The lichen-dominated vegetation types were changed into barrens, partly damaged vegetation entities, and dwarf shrub (e.g., Vaccinium myrtillus)-dominated vegetation. We found a significant negative relationship ( r=−0.94, p=0.001) between the extent of the area of mixed forests with lichen content and the emissions of SO 2 during the period. The area of the category industrial barrens had a significant negative relationship ( r=−0.95, p=0.001) to the emissions of SO 2 during the same period.

Validation of temperature-emissivity separation and split-window methods from TIMS data and ground measurements

Land surface temperature retrieved with temperature-emissivity separation (TES) and split-window (SW) algorithms from six-channel Thermal Infrared Multispectral Scanner (TIMS) data in the HAPEX-Sahel experiment agreed with contemporaneous ground temperature measurements to within ±1 °C (TES and SW with channels at 10.8 and 11.7 μm, or SW-56). The SW algorithm used with TIMS channels at 8.4 and 8.7 μm (SW-12) underestimated ground temperatures by 2–5 °C. The TES method required atmospheric correction of at-sensor radiances, which was done with local radiosonde data and MODTRAN 4, and an empirical relationship between the spectral range of emissivity and its minimum value. Emissivity data required for the SW algorithms were obtained using vegetation cover estimates from near-coincident reflective remote sensing data. The temperature underestimation of the SW-12 algorithm could be caused by errors in the emissivity inputs calculated from the vegetation cover. Such errors were due to the high variability of surface emissivity in the 8–9-μm waveband, which was much larger than in the 10–12-μm region. This was checked using TES derived emissivities as inputs of the SW algorithms, and comparing the resulting temperatures with the TES temperatures. In this case, both the SW-56 and SW-12 temperatures agreed with TES within ±1 °C for all sites and scenes.

Retrospective studies of grazing-induced land degradation: A case study in central Crete, Greece

Although it is well known that livestock husbandry has become more prevalent in Crete, Greece, during the last two decades, it is not well understood how grazing pressure differs spatially and how it influences grazed rangelands over time. To enhance the understanding of spatio-temporal degradation processes, a 20-year time series of Landsat Thematic Mapper and Multi-Spectral Scanner data is analysed. A set of vegetation fraction images is derived, based on a rigorous pre-processing scheme and linear spectral unmixing. Patterns of vegetation degradation, stability and regeneration are subsequently mapped by means of a trend analysis. It can be concluded that about 40% of the central Cretan rangelands show a declining trend in vegetation cover between 1977 and 1996. A deeper understanding of the underlying driving forces can be achieved through integrating remote sensing-derived information with other geoinformation sources. A comparison between local stress factors and vegetation development reveals that degradation processes are most evident when boundary conditions are favourable. It is, hence, assumed that increased grazing pressure is a major driving force for decreasing vegetation cover in the long term.

Remote sensing of agriculture change in Oman

This paper uses Landsat Multi-Spectral Scanner (MSS) data from 1978 and Landsat Enhanced Thematic Mapper (ETM) data from 2000 to assess agriculture change in two test areas of the Batinah coast of Oman. Field work was carried out in 1979 and 2001 in the Khaburah and Sohar regions to support the analysis of the Earth observation data. Unsupervised and supervised classifications are explored to produce maps of environmental change, and from these maps estimates of agriculture area are produced. While there has been substantial and clear agriculture change in the test areas, stimulated by Oman's rapid economic change since 1970, the measurement of these changes presents challenges that are common to many remote sensing investigations of arid and semi-arid regions. The paper identifies the main areas and types of agriculture change in the two test regions.

An approach for assessing boreal forest conditions based on combined use of satellite SAR and multi-spectral data

The degree and spatial distribution of boreal forest ecosystem degradation in Russia are not well known. The objective of this study is to develop an interpretation basis for analysis of satellite remote sensing data using a set of indicators characterizing the ecological situation and the degree of industrial pollution. European Remote Sensing Satellite (ERS) Synthetic Aperture Radar (SAR) and Landsat Multi-Spectral Scanner (MSS) data are used in combination for this purpose, along with an exceptionally extensive in situ data set of ground measurements of spectral radiance of pine biocenose components, and the results of moss chemistry and bio-indicator studies from the ecologically stressed St Petersburg region. It is shown that ERS SAR images provide an assessment of forested area distribution and forest type classification. The main factors of variability in parameters such as Normalized Difference Vegetation Index (NDVI) that are most strongly related to in situ indicators reflecting the state of the forest are identified. A supervised classification of forest degradation was performed on the basis of the NDVI values from the Landsat images. The results obtained make it possible to specify the areas at a local level and perform regional assessments. The potential for multi-temporal ERS SAR and multi-spectral sensor observations to trace the dynamics of changes in forest ecosystems is evaluated.

Spectroscopic study of the Moses Lake dune field, Washington: Determination of compositional distributions and source lithologies

Source lithologies and transport histories of materials within the Ephrata Fan are investigated. Data were collected using a variety of remote‐sensing, laboratory spectroscopic, and field observations and techniques. Laboratory thermal emission spectra were collected of bedrock within the Grand Coulee, dune samples, and clast deposits. Factor analysis, target transformation, and end‐member recovery techniques were applied to the set of dune samples as well as a set of grain size fractions. The dune sample spectra are composed of three components that represent basalt, granodiorite, and clay compositions. The basalt and granodiorite components are similar to spectra of clast and bedrock samples from the Grand Coulee and the Ephrata Fan. The clay component is similar to weathering surfaces located within the dune field. The same components were recovered from the set of grain size fractions from a single dune sample demonstrating a relatively higher basalt concentration with grain sizes greater than ∼250 μm. Thermal Infrared Multispectral Scanner (TIMS) data display significant intradune compositional variation and no discernable interdune compositional variation, indicating that the basalt and granodiorite components were likely deposited simultaneously and subsequently separated by wind based on grain size. Basalt and granodiorite bedrock units within the Channeled Scablands are source materials for the deposits within the Ephrata Fan and Moses Lake dune field. The Columbia River, located 20 km west of the dune field, is not a likely source of material.

Estimation of cirrus and stratus cloud height using landsat data and 3D graphical representation

Cirrus and stratus cloud heights are estimated using LANDSAT MSS(MultiSpectral Scanner) and TM(Thematic mapper) data. Cloud heights are calculated from the sun elevation angle and the horizontal distance between cloud and its shadow templates. The matching of these templates is accomplished by the 2D cross-correlation function. Before accomplishing the matching, histogram equalization, inversion of shadow template, and Fourier Fast Transformation are applied to both templates. Then 2D cross-correlation function is calculated in the frequency domain. Correlation coefficients are 3D graphically displayed to check visually how sharply the templates match. The effects of the skew angle and the nonnadir observation angle are taken into consideration to modify the direction from the cloud template to its corresponding shadow template. The results show that this method is accurate within several pixels.

A procedure for regional lake water clarity assessment using Landsat multispectral data

Although previous investigations have demonstrated reliable empirical relationships between satellite data and nearly contemporaneous ground observations, satellite imagery has not been incorporated into routine lake monitoring programs. This paper focuses on key issues involved in applying satellite imagery to the regional assessment of lake clarity. Ten Landsat Thematic Mapper (TM) images and four Multispectral Scanner (MSS) images of the Twin Cities Metropolitan Area (TCMA) spanning 25 years (1973–1998) were analyzed. Based on this analysis, recommendations are made for a Landsat-based procedure for water clarity assessment. Closeness of fit ( r 2 values) of regression models between satellite brightness data and measured Secchi disk transparency (SDT) decreased with increasing size of the time window between image collection and ground observation of SDT. Use of SDT data collected within ±1 day of the image date is recommended, but where SDT data are limited, windows up to ±7 days yield reasonable results, especially in late summer when water clarity is relatively constant. Average brightness data from at least nine pixels in the deep open area of a lake should be used to predict lake clarity, but the accuracy of predicted SDT did not improve much as the number of pixels in the area of interest (AOI) increased above this value. A three-coefficient regression model using the TM1/TM3 ratio and TM1 was a consistent and reliable predictor of SDT ( r 2 values of .7–.8). A similar relationship involving the MSS1/MSS2 ratio and MSS1 was a reasonable predictor of SDT for MSS data. Efforts to produce a standard prediction equation for SDT applicable to images collected on different dates were not successful, but a simple regression procedure to account for differences in atmospheric conditions among image collection dates substantially decreased the range of coefficients in the regression model.

Analysis of variations in land cover between 1972 and 1990, Kassala Province, Eastern Sudan, using Landsat MSS data

The paper describes an analysis of vegetation and land cover changes between 1972 and 1990 in Kassala Province, Eastern Sudan. Three areas were selected and studied using Landsat Multi-Spectral Scanner (MSS) data. A maximum likelihood routine was used and sand land, grassland, forest, irrigated areas, kerrib, water, cultivated areas and mountain areas were classified. The study indicates increases in sand land, forest and cultivated areas, and a decrease in grassland.

An estimate of forest cover extent and change in Jamaica using Landsat MSS data

Landsat Multispectral Scanner (MSS) data were used to derive an island-wide estimate of forest extent and change for Jamaica for the period 1987-1992. The study demonstrates how Landsat MSS data provide an invaluable resource for monitoring tropical forest change when, as in this study, a simple classification of forest versus non-forest is made. Using a broad definition of forest cover and a hybrid unsupervised/supervised classification, the study derived an average per annum island-wide deforestation rate of 3.9% for this period. This estimate was then compared to existing estimates for the island and its reasonabless argued for in view of the latter's limitations and the study's own accuracy assessment. By providing a broad snapshot of the rate of forest loss on the island, it is hoped that this study's country-wide estimate will highlight the severity of the deforestation problem to scientists and policy-makers, and will provide the basis for more detailed forest classification studies of the island.

Acacia canopy cover changes in Rawashda forest reserve, Kassala Province, Eastern Sudan, using linear regression NDVI models

The paper aims to assess the acacia canopy cover changes for the period 1972-1990 in the Rawashda forest reserve, Eastern Sudan. A linear regression model was applied using Landsat Multi-Spectral Scanner (MSS) data from 1972, 1979 and 1990. The study found an increase in canopy cover from 1972 to 1990, with the larger and denser forest areas occurring in 1990.

Ejecta distribution patterns at Meteor Crater, Arizona: On the applicability of lithologic end-member deconvolution for spaceborne thermal infrared data of Earth and Mars

[1] A spectral deconvolution using a constrained least squares approach was applied to airborne thermal infrared multispectral scanner (TIMS) data of Meteor Crater, Arizona. The three principal sedimentary units sampled by the impact were chosen as end-members, and their spectra were derived from the emissivity images. To validate previous estimates of the erosion of the near-rim ejecta, the model was used to identify the areal extent of the reworked material. The outputs of the algorithm reveal subtle mixing patterns in the ejecta, identified larger ejecta blocks, and were used to further constrain the volume of Coconino Sandstone present in the vicinity of the crater. The availability of the multialtitude data set also provided a means to examine the effects of resolution degradation and quantify the subsequent errors on the model. These data served as a test case for the use of image-derived lithologic end-members at various scales, which is critical for examining thermal infrared data of planetary surfaces. The model results indicate that the Coconino Ss. reworked ejecta is detectable over 3 km from the crater. This was confirmed by field sampling within the primary ejecta field and wind streak. The areal distribution patterns of this unit imply past erosion and subsequent sediment transport that was low to moderate compared with early studies and therefore places further constraints on the ejecta degradation of Meteor Crater. It also provides an important example of the analysis that can be performed on thermal infrared data currently being returned from Earth orbit and expected from Mars in 2002.

Remote sensing of snow grain-size and impurities from Airborne Multispectral Scanner data using a snow bidirectional reflectance distribution function model

AbstractAlgorithms to retrieve the snow grain-size and the concentration of snow impurities were developed using a theoretical bidirectional reflectance distribution function (BRDF) model of the snow surface. In this model, snow grains are assumed to be independent spherical ice particles, and the BRDF is calculated with multiple scattering by snow particles. Using these algorithms, the snow grain-size and snow impurities were retrieved from Airborne Multispectral Scanner (AMSS) images at the visible (λ = 0.545 μm) and near-infrared (λ =1.24,1.64 and 2.23 μm) wavelengths observed over the flat snowfield in eastern Hokkaido, Japan, in February 1998. The estimated snow grain-size and impurities were consistent with the results of in situ measurements on the snow surface. For snow grain-size, measured reflectances in the different near-infrared AMSS channels indicated grain-size differences in the vertical profile of the snowpack.

Recovery of spectral emissivity from Thermal Infrared Multispectral Scanner imagery acquired over a mountainous terrain: A case study from Mount Etna Sicily

The estimation of ground radiance and emissivity from Thermal Infrared Multispectral Scanner (TIMS) data is strongly dependent on the atmospheric correction applied to these data. Since such corrections are a function of the atmospheric path between the sensor and ground, correction techniques that do not consider the topographic variations within a scene can introduce appreciable error in the estimation of the atmosphere effects. In this paper, we describe the development and application of a variable-elevation atmospheric correction procedure. Our objective was to incorporate changes in target altitude into a general atmosphere correction strategy. This procedure was tested on a TIMS data set acquired over Mt. Etna, Italy in July 1986. The methodology adopted in this study is based on the use of the LOWTRAN radiative transfer code and a digital elevation model (DEM) registered to the image data. The image data are divided into a series of layers, based on elevation, and a separate atmosphere correction is applied to each layer. Maps of emissivity estimates derived with the variable-elevation approach were compared with geologic maps of the Etna flow fields. Prior to the variable-elevation correction, the emissivity spectra of long lava flows appeared to vary with elevation. Following the variable-elevation correction, many of these spectral artifacts were removed from the emissivity maps. In addition, the variable-elevation correction increased our ability to discriminate individual lava flows.

Effects of rainfall variability and communal and semi-commercial grazing on land cover in southern African rangelands

Semi-arid ecosystems in southern Africa are experiencing change due to natural vari- ability in rainfall and to changes in the type and intensity of land use. This paper applies grazing gra- dient techniques to vegetation cover dynamics in areas of communal and semi-commercial grazing in the Kalahari rangelands of Botswana to see whether human impact can be separated from that of climate variability. Landsat Thematic Mapper and Multispectral Scanner data were acquired for the Gumare and Hainaveld areas for a number of wet and dry seasons. Vegetation cover was mapped using the PD54 vegetation index. Gradients in both vegetation cover and vegetation response to rain- fall were identified under communal use, suggesting both human impact and human-induced degra- dation. However, patterns were complex because of multiple focal points in the hinterland of larger settlements. Gradients in vegetation response were more effective than gradients in cover as identi- fiers of human impact in semi-commercial grazing areas. These gradients intensify in dry periods but vegetation partially recovers after good rains. Over time, however, the gradients are intensifying, suggesting that long-term grazing impact is occurring. As gradients intensify and degradation increases, vulnerability of grazing systems to rainfall variability and to drought is likely to intensify.

Landscape cover dynamics pattern in Meghalaya

Human interventions in natural systems have resulted in large changes in vegetation composition and distribution patterns. The Land Use Change and Climate Change (LUCC) study under the International Geosphere Biosphere Program (IGBP) is a major initiative in this regard. Changes in land use and hence in vegetation cover, due to climatic change and human activity, affect surface water and energy budgets directly through plant transpiration, surface albedo, emissivity and roughness. They also affect primary production and, therefore, the carbon cycle. Thus, there is a need for spatial and temporal characterization of vegetation cover at different scales, from the global and continental scale to the local patch scale. Satellite remote sensing provides detailed information regarding the spatial distribution and extent of land use changes in the landscape. Meghalaya, in north-east India, is one of the most important, biologically rich landscapes. Degradational activities, namely shifting cultivation, clear felling of forests for timber, and mining, have altered the natural landscape to a great extent. Because of these increased anthropogenic activities the natural landscape has been modified which has resulted in a fragmented landscape with poor species composition. These changes in the landscape were analysed using IRS 1A, 1B and Landsat Multi-Spectral Scanner (MSS) data during the period 1980-1995. The vegetation type maps were prepared by a visual interpretation technique in order to study the land cover dynamics pattern in Meghalaya.

Assessment of environmental impacts of river basin development on the riverine forests of eastern Kenya using multi-temporal satellite data

The utility of Landsat MSS (Multispectral Scanner) and SPOT XS data in monitoring the impacts of river basin development on a riverine forest located in the lower Tana River Basin of eastern Kenya was evaluated. Land cover change maps derived from Landsat MSS indicated little change in total forest area between 1975 and 1984. Land cover change maps derived from SPOT XS data indicated a 27% decline in forest area between 1989 and 1996. Mean patch size and area-perimeter ratio of the closed riverine forest remained virtually unchanged whereas these parameters for the open forest class decreased by 31% and 4% respectively. In addition, the average extent of the open riverine forest from the river channel declined by about 200 m between 1989 and 1996. This decline was attributed to decreased extent of floods along the floodplain following construction of dams in the upper river basin, and increased exploitation of the forests for fuelwood, especially in the vicinity of the established Bura Irrigation and Settlement Project. The greater lateral movement observed in the location of the river channel for the 1975-1985 period, compared to the 1985-1996 period, was also attributed to construction of dams in the upper river basin.

The relations between spectral data and water in a crop production environment

The results of a correlation study between high spatial resolution (1m) Multi-Spectral Scanner (MSS) data and reference data with particular emphasis to soil-plant water parameters are presented. Digital images of the Management Systems Evaluation Area (MSEA) research site in south central Ohio were acquired using MSS mounted on an aircraft. Gravimetric soil water, plant water, and plant residue information relating to MSS data were collected in April, July, August, and September, 1994. It was found that the correlation between MSS data and soil water was changed by the presence of crop residue versus bare soil. The combined water estimates from the soil and plant enhanced the correlation structure between MSS parameters and soil-plant water variables. It was also shown that when using soil water content data below the 0.5-2.0 cm soil depth, the correlation with MSS parameters decreased.