• About
  • Advanced Search
  • Browse Proceedings
  • Access Policy
  • Sponsor
  • Contact
  • Development in applications of remote sensing to hydrology

    Paper ID



    • Andrew Adelman
    • Reuben Ambaruch
    • John W. Simmons


    IBM, Federal Systems Division, Electronics Systems Center






    The progress and initial results in the first major part of a streamflow forecasting project, utilizing aircraft (and later space) derived imagery for watersheds of the Tennessee River Valley, are presented. The initial study phase, aimed at assessing the feasibility of applying remote sensed data to prediction of watershed performance, is approximately thirty percent complete. Demonstration of feasibility will lead to a means of prediction of the hydrological behavior of ungaged watersheds using remote sensed data to minimize time, effort, and cost of achieving these predictions, as well as provide good prediction capability with fewer ground instruments in existing installations. Catchments have been chosen for calibration and simulation from the Tennessee Valley region, for which a wealth of climatological and streamflow historical data as well as extensive aerial photographic and ground survey coverage are available. An initial set of 14 catchments chosen for analysis represents five distinct physiographic provinces in the region and vary in size from 6.9 to 365 km2. A computer program based on the well-known Stanford Watershed Model IV and a companion para- meter-optimization program are used to calibrate model parameters for the selected catchments, based on historical data. The accuracy of the models is shown by comparison of synthesized with observed stream- flows. Methods of determining model parameters from physical characteristics, observable or inferable from remotely acquired photographs, without the necessity of historical streamflow data, are discussed. Extension of the effort includes validating applicability of the models to catchments in other regions of the world. Sensitivity of parameter estimation and simulation accuracies to lower resolution in sensor data, such as Earth observation data acquired from space, will be assessed. Detailed procedures for applying the models will be prepared and made available for potential use in all parts of the world.