Gary W. Petersen

Texture transforms of remote sensing data

Abstract The application of image texture to the digital analysis of remotely sensed data requires a quantitative characterization of texture. Several methods reported in the literature characterize the spatial distribution of gray levels across an image segment as texture measures. These texture measures represent all the pixels within an image segment as samples from the same population. Another approach to texture analysis is the construction of texture transforms consisting of discrete measures for each pixel. This technique describes the gray-level distributions around each pixel. This paper reports the analysis of texture transforms of multispectral digital imagery. The transforms were derived from local properties computed within three-by-three pixel windows surrounding each pixel. The local properties were computed using a combination of available channels. The properties included: the mean, variance, skewness, and kurtosis of gray levels; mean and maximum gray-level differences; and mean and maximum Euclidean distances between gray level vectors. Logarithmic transformations of the local properties were used to expand the contrast of the texture transforms. Texture transforms of Landsat-2 MSS data were generated. The transforms were useful for edge detection and image enhancement, but did not prove useful as features for the thematic mapping of land cover. These results are compared to a previous investigation for which texture transforms of digitized aerial photography were effectively employed as features for the thematic mapping of terrain categories.

Environmental Indicators of Agroecosystems

Conventional production agricultural practices are partly responsible for intensifying the degradation of productive lands throughout the world. In monitoring the impacts of these practices, a variety of biological, physical, chemical, landscape, and economic measures are being used as indicators of environmental change. This chapter is largely a review of both common and uncommonly used environmental indicators of agricultural systems. Soil organic matter content is discussed in detail as a candidate environmental indicator. A ranking scheme is proposed for the use of multiple indicators in decision-making applications.

Geographic Information Systems in Agronomy

Publisher Summary This chapter is intended to provide an introduction to geographic information system (GIS) and associated landscape tools and to illustrate the ways in which they are being used in various aspects of agronomy. GIS technology is bringing about rapid changes in the way that agronomic analysis and management are being conducted. GIS coupled with remote sensing, Global Positioning System (GPS), electronic sensors, and computer technologies is providing new methods for data acquisition, storage, processing, analysis, and modeling. These new tools allow us to quantitatively describe landscapes and processes. The chapter discusses site-specific farming (SSF)—that is, farm management based upon variable soil and microclimate conditions that occur within most fields. SSF reduces waste, because fertilizer and herbicide—for example—are applied only where needed. New and/or improved models need to be developed to fully take advantage of the spatial nature of the data provided by these tools. The development of these models will rely on spatial statistical analysis techniques to quantify the accuracy of input parameters and model output. Many new tools are being used in this rapidly evolving field of GIS. Three-dimensional scene simulation, visualization, and animation linked with remote sensing and image processing technologies, and real time data collection will be needed in the study of agronomic systems. The development and use of three-dimensional GIS and spatiotemporal GIS will be an increasingly important area of research. The agronomic community—including farmers, land managers, fellow scientists, policymakers, and the general public should benefit from this evolving and expanding field.

A WebGIS system for relating genetic soil classification of China to soil taxonomy

Soil classification is the basis for the exchange of soil science research results and the foundation for the application of modern soil resource management methods. A WebGIS-based system designed to relate genetic soil classification of China (GSCC) to soil taxonomy (ST) was developed to enhance global cooperation and to support communication between China and the other countries on important agricultural and environmental issues. The system has a Browse Server (B/S) structure and exploits the 1:1,000,000 soil databases of China using WebGIS functionality. This paper describes the application of the WebGIS system for easily accessing cross-reference information between GSCC to ST. First, we describe the three-level B/S structure of the system. The cross-reference methodologies, referenceability and maximum referenceability, are then explained and applied at three geographic scales (i.e. nation, region and pedon). Finally, three sub-modules based on the supported scales are described and illustrated with application scenarios to familiarize users with the inquiry system and its usage. The main advantage of the system is that it considers statistical similarity in the spatial distributions between the two different classification systems. Users with limited knowledge are able to obtain soil cross-reference information using an intuitive interface, which supports query, visualization and analysis via a web browser at the most detailed level. The inquiry system benefits the development of soil classification science and international academic exchange.

Aircraft and satellite remote sensing of desert soils and landscapes

Abstract The remote sensing of desert soils and landscapes using Thematic Mapper (TM), Heat Capacity Mapping Mission (HCMM), Simulated SPOT, and Thermal Infrared Multispectral Scanner (TIMS) data is discussed. These studies were all conducted in arid or semiarid study sites. Landsat Thematic Mapper (TM) data for southwestern Nevada discriminated among alluvial fan deposits with different degrees of desert pavement and varnish as well as different vegetation cover. Thermal-infrared data acquired from the Heat Capacity Mapping Mission (HCMM) satellite were used to map the spatial distribution of diurnal surface temperatures and to estimate mean annual soil temperatures in semiarid east central Utah using diurnal data for five dates throughout a year. Simulated SPOT data for northwestern New Mexico identified geomorphic features, such as differences in eolian sand cover and fluvial incision, which are correlated with surface age and geomorphic stability of landscape components. The Thermal Infrared Multispectral Scanner (TIMS), which is an aircraft scanner that provides six-channel spectral capability in the thermal region of the electromagnetic spectrum, was used to depict surface geologic features of the Saline Valley in southeastern California. These research projects are presented as a summary of some of the sensors and analytical techniques that are useful in the study of desert soils and landscapes.

Spectral signature selection for mapping unvegetated soils

Abstract Airborne multispectral scanner data covering the wavelength interval from 0.40–2.60 μm were collected at an altitude of 1000 m above the terrain in southeastern Pennsylvania. Uniform training areas were selected within three sites from this flightline. These training areas were field mapped as Berks, Duffield and Penn soils. Soil samples were collected from each site and a procedure developed to allow assignment of scan line and element number from the multispectral scanner data to each sampling location. These soil samples were analyzed on a Beckman DK-2A Spectrophotometer and laboratory spectral signatures derived. After correcting for solar radiation and atmospheric attenuation, these laboratory signatures were compared to the spectral signatures derived from these same soils using multispectral scanner data. Both signatures were used in supervised and unsupervised classification routines. Computer-generated maps using the laboratory and multispectral scanner derived signatures resulted in maps that were similar to maps resulting from field surveys. Approximately 90% agreement was obtained between classification maps produced using multispectral scanner derived signatures and laboratory derived signatures.

C‐band, multi‐angle SAR imaging of agricultural cover

The launch of the Canadian Space Agency’s RADARSAT will provide the capability to supply imagery acquired from different instrument look angles, resulting in a multi‐incidence view of an area of interest. Previous research has utilized multi‐incidence imaging to map land use/land cover and to model soil moisture content. This research utilizes an innovative approach to the use of multi‐angle SAR imagery to detect differences in land use/land cover. In this study, plant canopies and bare soil are imaged with multi‐angle and multi‐frequency data acquired with the airborne NASA/JPL AIRSAR instrument. Findings are relevant for possible applications of data from the soon to be orbited RADARSAT instrument. Backscatter from a distributed target, such as vegetation, is the product of microwave absorption and scattering interactions with all features within the imaged area. The intensity of the backscatter from the area varies with the angle of the incident wave. This investigation uses this physical understanding...

An assessment of soil productivity loss caused by expanding urban land use using remote sensing and soil productivity models

An EOS IDS project has been recently designed to assess the loss of soil productivity resulting from expanding urbanization in the U.S. and selected regions in Mexico and the Middle East using remotely sensed data and soil productivity models. The extent of urbanization will be determined by generating urban land cover layers from DMSP/OLS (Defense Meteorological Satellite Program’s Operational Linescan System) nighttime imagery. This imagery will be calibrated using Landsat Thematic Mapper (TM) and population/housing census data. A range of soil/land productivity models will be evaluated using soil factors computed from the State Soil Geographic Database (STATSGO) and FAO soil databases, terrain models, climate and vegetation to rank soil mapping units based on their productivity potential. Examples of these models are the Net Primary Productivity (NPP) and FAO Fertility Capability Classification (FCC) system. The magnitude of soil productivity loss due to urbanization will finally be determined by analysis of data obtained from GIS overlays of urban land use and soil productivity layers.