Tactical Mobility Modeling for REFORGER 87
Total Page:16
File Type:pdf, Size:1020Kb
Tactical Mobility Modeling for REFORGER 87 J. David Lashlee and James H. Robinson U. S. Army Engineer Waterways Experiment Station, Vicksburg, MS 39180-6199 ABSTRACT: The Waterways Experiment Station (WES) responded to requests by the U.s. Army III Corps, 64th Engineer Detachment (Terrain) by providing trafficability support for the REFORGER 87 military exercises in the Federal Republic of Germany. The maneuver area covered 25 1:50,000-scale topographic map sheets. A digital mobility database was developed within a period of one month. A principal source of information was Landsat Thematic Mapper (TM) data, which was integrated with various other digital factor maps necessary for computer modeling of vehicle mobility. The Condensed Army Mobility Model System (CAMMS) was used to evaluate the performance of specified vehicles operating off-road in the maneuver area. CAMMS correctly predicted a NOGO situation for an area where several U.s. armored vehicles were immobilized when a gravel road failed and the vehicles were unable to traverse the surrounding organic clay soil and associated drainage ditches. \r __ " \ I INTRODUCTION I a." -'f(' '-- BACKGROUND ,Y ""p QUAD CODE NO. :=:..tI ? ---- -- ,r::.c----r-- { h. 1 2 720 In January 1987, the U.S. Army III Corps, 64th Engineer De - 2 2722 r I 3 2 72. tachment, began planning for the 1987 REFORGER (Return of lJ" 2726 1 12 3 4 . 5!'-- 5 2728 Forces to Germany) military exercises to be held in September 6 7 8 9 , 2920 I 10 I\..---. 7 2'22 of that year. On 30 June 1987, a formal request for WES's support "( 12 13 14 2,,, 151 I 111 Jr-V ,• 2'26 , I 16 17 18 19 201 "- 10 292. for mobility analyses was made by III Corps. The area of op ..-J 11 3120 l.... I 2123 24 25 12 3122 erations consisted of 25 1:50,000-scale topographic maps located -r--- \ 13 3124 ~I-- J~ 1. 3126 in the State of Lower Saxony. It was requested that WES de 15 3128 .( J l' 3320 velop a digital mobility database covering the exercise area and ~.,.-' 17 3322 1. 3324 provide the necessary hardware, software, and technical sup 3326 I'20 3328 port for trafficability analyses in the field. Due to the size of the "- 21 3520 22 3522 -l 23 3524 study area and time constraints, it was decided that land-cover 3 '26 delineation by manual cartographic and digitization techniques r/ ~ "25 3528 was impractical. However, derivation of land-cover information L \. I ( from digital satellite data was feasible. Landsat Thematic Map ( per data are the preferred data source for most mobility and d I '- l ('f1P-=:1 combat modeling applications at WES (Lashlee and Overton, [ I I \. 1991). Only upon confirmation of the existence of appropriate \J I I \ \) 1M data was the decision made to develop the digital data base I I 1\ (Figure 1). A project methodology and time schedule were de C. I \ I \ I \ veloped and data processing began on 5 August 1987. II '--" ( I \ -1b THE RETURN OF FORCES TO GERMANY I I I I I I I I lI> f--f-- FEDERAL REPUBLIC OF GERMANY I lJ The United States has joined NATO for the annual REFORGER f I I \ I II V military exercises since 1969. REFORGER 87 was designed to em ( I I\\ \( phasize the U.S. capability to transport an entire Corps of com I I I I \ ~ bat forces for rapid reinforcement of Europe. Most of these ( ~ I I IA\] deploying units utilize equipment and material prepositioned rvJr' in Europe as part of NATO strategy. The purpose of these ex 'z:; ercises is to deter war in Europe. The exercise showed visiting FIG. 1. WES mobility database and Landsat TM data coverage. communist officials that NATO Doctrine is defensive in nature (Cox, 1988). Exercise CERTAIN STRIKE, the ten-day battle phase of REFOR and 250 Howitzers and Multiple Launched rocket systems were GER 87, began on 14 September 1987 and simulated a Soviet and deployed to the 400 square kilometre maneuver box. REFORGER Communist East Block offensive thrust across the inner-German 87 was the largest deployment of U.s. forces since World War border, into the northern plains of the Federal Republic of Ger II (Cox, 1988). many. A total of 115,000 NATO military personnel from six NATO nations, including 35,000 U.S. combat soldiers, participated in STUDY AREA the exercise. Other participating NATO countries and approxi mate military personnel were the Federal Republic of Germany The North German Plain, one of four physiographic prov (19,000), The Netherlands (11,100), United Kingdom (7,000), inces, covers approximately one-third of the Federal Republic Belgium (6,000), and France (200). Twenty-thousand wheeled of Germany and is characterized by landforms that have de vehicles, 2,200 light armored vehicles, 700 main battle tanks, veloped almost exclusively in glacial deposits. It is primarily a PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 0099-1112/92/5808-1223$03.00/0 Vol. 58, No.8, August 1992, pp. 1223-1227. ©1992 American Society for Photogrammetry and Remote Sensing 1224 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING, 1992 region of moraines and glacio-fluvial sediments (Hagedorn, 1989). The SMSP makes predictions based on historical precipitation Soils in this area are generally coarse grained sands and silts data, precipitation measured for the prior 24-hour period, and with isolated occurrences of fine grained and organic soils. from the current soil moisture content adjusted by the fore Drainage in the area consists of the Aller, Wesser, and Elbe casted precipitation for multiples of 24 hours (Butler, 1988). Rivers, and their tributaries, which drain into the North Sea. The topography is generally flat to rolling with local differences in elevation between 76 and 152 metres (250 and 500 feet) above DIGITAL MOBILITY DATA BASE DEVELOPMENT mean sea level. Slopes are usually less than 10 percent, but can DIGITAL IMAGE ANALYSIS be as steep as 45 percent on isolated landforms. The North German Plain generally affords good cross country movement TM bands 2,3,4, and 5 were chosen for image processing. The during the dry months. However, numerous marshlands, peat software used to process these data was the Earth Resources bogs, canals, small streams, and an extensive network of drain Laboratory Applications Software (ELAS), developed by the Na age ditches are common obstacles to off-road mobility. The tac tional Aeronautics and Space Administration (NASA) (Graham tical exercise box covered 400 square kilometrers, ranging from et aI., 1980). Thirty-six unsupervised training statistics were cal the city of Hannover, Lower Saxony's Capital city, to the Elbe culated with a window based clustering algorithm. The defi River in the Northeast. Lower Saxony has a population of 7.3 nition of a homogeneous training field is a 3 by 3 window with million. a standard deviation between user selected homogeneity pa rameters for the standard deviation lower bound (Default = MOBILITY MODELING 0.1) and either the standard deviation upper bound (Default = AMM, NRMM, AND CAMMS 1.0) or the coefficient of variation (Default = 5 percent) times the mean of the window values. Each statistic accepted by the Mobility models are comprehensive analysis tools designed program consists of the mean, standard deviation, and covari to objectively evaluate the on- and off-road mobility of vehicles ance of a homogeneous 3 by 3 window of pixels. by means of digital computer simulation. Military ground ve Mter the initial statistics were calculated, a procedure was hicle field testing and mathematical modeling have been con executed to merge similar candidate statistics into a managea ducted at WES since World War II. Results of these tests and ble, user defined number of final statistics, thus insuring a min models were consolidated in a computerized mobility model in imum statistical scaled distance between the means of the 1971. The Army Mobility Model (AMM) predicts vehicle traffic statistics. The scaled distance equation is essentially the square ability as a function of the vehicle, driver, terrain, weather, and root of the sum of squares of the differences in the means be scenario factors that influence on-road, off-road, and gap cross tween a pair of statistics, normalized by the variation in each ing vehicle mobility. The AMM software has undergone exten band at the merged mean. sive field testing and validation. In 1978, the AMM was accepted To maximize the overall classification accuracy, a Bayesian as Edition 1 of the NATO Reference Mobility Model (NRMM). classification algorithm was used to assign each pixel to a land Internationally accepted, the NRMM includes vehicle mobility cover category. The ELAS Bayesian classifier uses the a priori prediction relations that are modified only by formal consensus probabilities determined from the number of pixels collected in of the NATO countries that use the NRMM (Turnage, 1989). The each statistic in an equation to determine the probability that a AMM and NRMM have been used extensively in U.s. military given pixel vector X belongs to a class. vehicle acquisition programs. Ground photographs, orthophotographic maps, German pro The Condensed Army Mobility Model System (CAMMS) was duced military topographic maps, engineering geology maps, developed by WES in 1983. A derivative of the AMM, CAMMS is statistical field data, and previous field experience of WES per a vehicle mobility analysis model designed for U.S. Army tac sonnel were used to assign information classes to the 36 spectral tical applications. CAMMS provides a comprehensive description classes. The land-cover image was georeferenced to a Universal of the ability of vehicles and vehicle convoys to transport men Transverse Mercator map projection and resampled to 100-metre and material over virtually any type of terrain and under nearly spatial resolution for compatibility with other factor maps de any weather condition (Butler, 1988).