Search

The vulnerability of dryland surfaces to wind erosion depends importantly on the absence or the presence and character of surface roughness elements, such as plants, clasts, and topographic irregularities that diminish wind speed near the surface. A model for the friction velocity ratio has been developed to account for wind sheltering by many different types of co-existing roughness elements. Such conditions typify a monitored area in the central Mojave Desert, California, that experiences frequent sand movement and dust emission. Two additional models are used to convert the friction velocity ratio to the surface roughness length (zo) for momentum. To calculate roughness lengths from these models, measurements were made at 11 sites within the monitored area to characterize the surface roughness element. Measurements included (1) the number of roughness species (e.g., plants, small-scale topography, clasts), and their associated heights and widths, (2) spacing among species, and (3) vegetation porosity (a measurement of the spatial distribution of woody elements of a plant). Documented or estimated values of drag coefficients for different species were included in the modeling. At these sites, wind-speed profiles were measured during periods of neutral atmospheric stability using three 9-m towers with three or four calibrated anemometers on each. Modeled roughness lengths show a close correspondence (correlation coefficient, 0.84–0.86) to the aerodynamically determined values at the field sites. The geometric properties of the roughness elements in the model are amenable to measurement at much higher temporal and spatial resolutions using remote-sensing techniques than can be accomplished through laborious ground-based methods. A remote-sensing approach to acquire values of the modeled roughness length is particularly important for the development of linked surface/atmosphere wind-erosion models sensitive to climate variability and land-use changes in areas such as the southwestern United States, where surface roughness has large spatial and temporal variations.

Sidescan-sonar imaging systems with digital capabilities have now been available for approximately 20 years. In this paper we present several of the various digital image processing techniques developed by the U.S. Geological Survey (USGS) and used to apply intensity/radiometric and geometric corrections, as well as enhance and digitally mosaic, sidescan-sonar images of the Monterey Bay region. New software run by a WWW server was designed and implemented to allow very large image data sets, such as the digital mosaic, to be easily viewed interactively, including the ability to roam throughout the digital mosaic at the web site in either compressed or full 1-m resolution. The processing is separated into the two different stages: preprocessing and information extraction. In the preprocessing stage, sensor-specific algorithms are applied to correct for both geometric and intensity/radiometric distortions introduced by the sensor. This is followed by digital mosaicking of the track-line strips into quadrangle format which can be used as input to either visual or digital image analysis and interpretation. An automatic seam removal procedure was used in combination with an interactive digital feathering/stenciling procedure to help minimize tone or seam matching problems between image strips from adjacent track-lines. The sidescan-sonar image processing package is part of the USGS Mini Image Processing System (MIPS) and has been designed to process data collected by any ‘generic’ digital sidescan-sonar imaging system. The USGS MIPS software, developed over the last 20 years as a public domain package, is available on the WWW at: http://terraweb.wr.usgs.gov/trs/software.html .

As part of a joint Russian/American dust-storm experiment, GOES-VISSR (Geostationary Operational Environmental Satellite, Visible-Infrared Spin-Scan Radiometer), data from a visible-band satellite image of a large dust storm emanating from Owens Lake, California were acquired on March 10 and 11, 1993. The satellite data were calibrated to targets of known ground reflectance factors and processed with radiative transfer techniques to yield aerosol (dust) optical depth at those stages of the dust storm when concurrent ground-based measurements of optical depth were made. Calibration of the satellite data is crucial for comparing surficial changes in remotely sensed data acquired over a period of time from the same area and for determining accurate concentrations of atmospheric aerosols using radiative transfer techniques. The calibration procedure forces the distribution of visible-band, DN (digital number) values, acquired on July 1, 1992, at 1731 GMT from the GOES-VISSR sensor over a large test area, to match the distribution of visible-band, DN values concurrently acquired from a Landsat MSS (Multispectral Scanner) sensor over the same test area; the Landsat MSS DN values were directly associated with reflectance factors measured from ground targets. The calibrated GOES-VISSR data for July 1, 1992, were then used to calibrate other GOES-VISSR data acquired on March 10 and 11, 1993, during the dust storm. Uncertainties in location of ground targets, bi-directional reflectance and atmospheric attenuation contribute an error of approximately ±0.02 in the satellite-inferred ground reflectance factors. On March 11 at 1031 PST the satellite-received radiances during the peak of the storm were 3 times larger than predicted by our radiative transfer model for a pure clay dust plume of infinite optical depth. This result supported ground-based measurements that the plume at that time was composed primarily of large salt grains, probably sodium sulfate, which could not be properly characterized in our radiative transfer model. Further, the satellite data showed that the salt fell out of the plume within 35 km from the source. Finer-grained, clay dust was observed to extend beyond the salt-laden plume and was the major component of the dust plume after 1131 PST, when erosion of the salt crust on Owens Lake ceased. By 1331 and 1401 PST satellite-inferred, optical depths compared favorably with measurements concurrently acquired at the ground. Uncertainties in bi-directional reflectance, atmospheric attenuation, and locating ground points in the satellite data manifest errors between the inferred and measured optical depths in the range of 20 to 50%; these errors would be much greater without the calibration of the GOES-VISSR data. Changes in satellite-inferred reflectance factors over the lake bed during the course of the storm showed that 76 km 2 of the surface was disrupted during the March 11 storm, suggesting as much as 76 × 10 3 m 3 of crustal material were displaced for each millimeter of several estimated to have been moved during the storm; an unknown fraction of the displaced material was suspended. The satellite data also showed dust fallout on mountain snowfields. Whereas fallout may have removed most of the salt, satellite data acquired at 1631 PST, when the plume had a large brightness contrast with the ground, showed that it covered over 2500 km 2 and contained at least 1.6 × 10 9 g of sediment. For such a small source area, the dust represents a substantial contribution to the regional and global load of aerosols.

The availability and use of remotely sensed digital image data have increased rapidly over the last two decades. One application where remotely sensed image data are a critical component is environmental analysis and monitoring. In our case it involves using sidescan sonar images to detect and map the location of 55‐gal drums that were disposed of near the Farallon Islands (about 50 nautical miles west of San Francisco). About 48,000 drums containing low‐level radioactive waste were dumped in water depths ranging from 100 to 2500 m covering 1400 km2. The high number of drums littering the seafloor and the rather large area involved makes remotely sensed image data one of the few promising tools to detect and map drum locations. In our analysis, sidescan sonar images collected by the SeaMarc IA and having about 1‐m pixel resolution were used. Spatial variability analysis (SVA) and other digital processing procedures were applied to the sidescan sonar images to automatically detect and map the location of t...

RESUMEPar l'analyse visuelle d'images de teledetection, on peut constater que celles-ci renferment de l'information spectrale et de l'information spatiale. La majorite des donnees de teledetection acquises ces vingt dernieres annees sont des donnees multibandes, et la plupart des methodes d'analyse utilisent davantage l'information spectrale que l'information spatiale. Toutefois, certains systemes (comme les sonars a balayage lateral ou les radars imageurs) recueillent les donnees dans une seule bande; l'utilisation de l'information spatiale contenue dans ces ensembles de donnees est donc particulierement importante.Cet article presente une methode destinee a extraire l'information sur l'amplitude et la variabilite spatiales de donnees numeriques obtenues par teledetection. Des filtres passe-haut sont utilises pour calculer l'indice d'amplitude spatiale et l'indice de variabilite spatiale aux echelles locale, intermediaire et regionale. Utilises comme donnees d'entree pour l'analyse des composantes princi...

Digital image processing and integration of data sets have been used to develop exploration models from airborne electromagnetics (EM), magnetics, and very‐low‐frequency electromagnetics (VLF-EM) data collected over an area in northwestern Arizona. The area has potential for the occurrence of uranium‐mineralized breccia pipes. Apparent resistivity and overburden thickness were derived from the EM measurements using half‐space models. Digital image processing techniques applied to the geophysical data sets included: (1) conversion of the data into gridded‐scale images, (2) spatial filtering for noise removal, (3) integration and analysis of the data sets, and (4) modeling using various parameter combinations. The general relationships between the geophysical variables/parameters and their ability to detect metallic deposits were used as guides in selecting digital number ranges that were used as input into various models. One of the best models incorporated apparent resistivity and total‐field magnetics; the results of this model outlined 13 anomalous combinations in the survey area. Field checking confirmed that two of the anomalies were previously known orebodies, and most of the other anomalies corresponded to suspected pipes that were under evaluation by the group that is exploring the property.

Digital analysis of remotely sensed data has become an important component of many earth-science studies. These data are often processed through a set of preprocessing or “clean-up” routines that includes a correction for atmospheric scattering, often called haze. Various methods to correct or remove the additive haze component have been developed, including the widely used dark-object subtraction technique. A problem with most of these methods is that the haze values for each spectral band are selected independently. This can create problems because atmospheric scattering is highly wavelength-dependent in the visible part of the electromagnetic spectrum and the scattering values are correlated with each other. Therefore, multispectral data such as from the Landsat Thematic Mapper and Multispectral Scanner must be corrected with haze values that are spectral band dependent. An improved dark-object subtraction technique is demonstrated that allows the user to select a relative atmospheric scattering model to predict the haze values for all the spectral bands from a selected starting band haze value. The improved method normalizes the predicted haze values for the different gain and offset parameters used by the imaging system. Examples of haze value differences between the old and improved methods for Thematic Mapper Bands 1, 2, 3, 4, 5, and 7 are 40.0, 13.0, 12.0, 8.0, 5.0, and 2.0 vs. 40.0, 13.2, 8.9, 4.9, 16.7, and 3.3, respectively, using a relative scattering model of a clear atmosphere. In one Landsat multispectral scanner image the haze value differences for Bands 4, 5, 6, and 7 were 30.0, 50.0, 50.0, and 40.0 for the old method vs. 30.0, 34.4, 43.6, and 6.4 for the new method using a relative scattering model of a hazy atmosphere.

Edge enhancement is a technique that can be considered, to a first order, a correction for the modulation transfer function of an imaging system. Digital imaging systems sample a continuous function at discrete intervals so that high-frequency information cannot be recorded at the same precision as lower frequency data. Because of this, fine detail or edge information in digital images is lost. Spatial filtering techniques can be used to enhance the fine detail information that does exist in the digital image, but the filter size is dependent on the type of area being processed. A technique has been developed by the authors that uses the horizontal first difference to automatically select the optimum kernel-size that should be used to enhance the edges that are contained in the image.

The Landsat albedo, or percentage of incoming radiation reflected from the ground in the wavelength range of 0.5 μm to 1.1 μm, is calculated from an equation using the Landsat digital brightness values and solar irradiance values, and correcting for atmospheric scattering, multispectral scanner calibration, and sun angle. The albedo calculated for each pixel is used to create an albedo image, whose grey scale is proportional to the albedo. Differencing sequential registered images and mapping selected values of the difference is used to create quantitative maps of increased or decreased albedo values of the terrain. All maps and other output products are in black and white rather than color, thus making the method quite economical. Decreases of albedo in arid regions may indicate improvement of land quality; increases may indicate degradation. Tests of the albedo difference mapping method in the Desert Experimental Range in southwestern Utah (a cold desert with little long-term terrain change) for a four-year period show that mapped changes can be correlated with erosion from flash floods, increased or decreased soil moisture, and increases or decreases in the density of desert vegetation, both perennial shrubs and annual plants. All terrain changes identified in this test were related to variations in precipitation. Although further tests of this method in hot deserts showing severe “desertification” are needed, the method is nevertheless recommended for experimental use in monitoring terrain change in other arid and semiarid regions of the world.