Search USGS⌕ Search

Geology topics

Wayne G. Rohde

Publications and source records attributed to Wayne G. Rohde.

10 recordsLinked to original sources
USGS1979

Inventory and mapping of flood inundation using interactive digital image analysis techniques

LANDSAT digital data and color infra-red photographs were used in a multiphase sampling scheme to estimate the area of agricultural land affected by a flood. The LANDSAT data were classified with a maximum likelihood algorithm. Stratification of the LANDSAT data, prior to classification, greatly reduced misclassification errors. The classification results were used to prepare a map overlay showing the areal extent of flooding. These data also provided statistics required to estimate sample size in a two phase sampling scheme, and provided quick, accurate estimates of areas flooded for the first phase. The measurements made in the second phase, based on ground data and photo-interpretation, were used with two phase sampling statistics to estimate the area of agricultural land affected by flooding These results show that LANDSAT digital data can be used to prepare map overlays showing the extent of flooding on agricultural land and, with two phase sampling procedures, can provide acreage estimates with sampling errors of about 5 percent. This procedure provides a technique for rapidly assessing the areal extent of flood conditions on agricultural land and would provide a basis for designing a sampling framework to estimate the impact of flooding on crop production.

Canadian Journal of Remote Sensing↗
USGS1978

EROS Data Center Landsat digital enhancement techniques and imagery availability

The US Geological Survey's EROS Data Center (EDC) is experimenting with the production of digitally enhanced Landsat imagery. Advanced digital image processing techniques are used to perform geometric and radiometric corrections and to perform contrast and edge enhancements. The enhanced image product is produced from digitally preprocessed Landsat computer compatible tapes (CCTs) on a laser beam film recording system. Landsat CCT data have several geometric distortions which are corrected when NASA produces the standard film products. When producing film images from CCT's, geometric correction of the data is required. The EDC Digital Image Enhancement System (EDIES) compensates for geometric distortions introduced by Earth's rotation, variable line length, non-uniform mirror scan velocity, and detector misregistration. Radiometric anomalies such as bad data lines and striping are common to many Landsat film products and are also in the CCT data. Bad data lines or line segments with more than 150 contiguous bad pixels are corrected by inserting data from the previous line in place of the bad data. Striping, caused by variations in detector gain and offset, is removed with a destriping algorithm applied after digitally enhancing the data. Image enhancement is performed by applying a linear contrast stretch and an edge enhancement algorithm. The linear contrast enhancement algorithm is designed to expand digitally the full range of useful data recorded on the CCT over the range of 256 digital counts. This minimizes the effect of atmospheric scattering and saturates the relative brightness of highly reflecting features such as clouds or snow. It is the intent that no meaningful terrain data are eliminated by the digital processing. The edge enhancement algorithm is designed to enhance boundaries between terrain features that exhibit subtle differences in brightness values along edges of features. After the digital data have been processed, data for each Landsat band are recorded on black-and-white film with a laser beam film recorder (LBR). The LBR corrects for aspect ratio distortions as the digital data are recorded on the recording film over a preselected density range. Positive transparencies of MSS bands 4, 5, and 7 produced by the LBR are used to make color composite transparencies. Color film positives are made photographically from first generation black-and-white products generated on the LBR.

Canadian Journal of Remote Sensing↗
USGS1978

Potential applications of satellite imagery in some types of natural resource inventories

Landsat satellite imagery has been routinely available to researchers and resource scientists since 1972. Many studies have demonstrated the application of Landsat imagery for conducting inventories and mapping various natural resources. Examples of applications presented in this paper include: timber volume inventory, range productivity inventory, wildland vegetation mapping, inventory of rangeland conversion and irrigated lands, and mapping strip mine disturbance in forested areas. Costs and accuracies for each of these application areas are presented. Landsat data, used in its proper context with appropriate analysis techniques, supporting data, and sound sampling strategies, can be an effective tool in conducting natural resource inventories. Other Landsat systems, scheduled for launch in 1978 and 1981, will provide continuity in the availability of Landsat data to users. Improvements in the quality and timeliness of data delivered to users, along with improved data handling procedures, analysis techniques, and improved sampling frameworks should increase the use of satellite imagery in natural resource inventories.

Conference Paper↗
USGS1977

A selected bibliography: Application of Landsat digital multispectral scanner data to agriculture, forestry, and range management

This bibliography contains citations of selected publications and technical reports dealing with the application of Landsat digital data analysis techniques to agriculture, forestry, and range management problems. All of the citations were published between 1973 and 1977. The citations reference publications and reports which discuss specific analysis techniques and specific resource applications.

Report↗