skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Automated Approach for Construction of Long-Term, Data-Intensive Watershed Models

Abstract

Watershed models such as the Hydrological Simulation Program in FORTRAN (HSPF) are frequently used to analyze large-scale water quantity and water quality issues. The construction of HSPF models is a difficult and time-consuming process, because it requires compilation of extensive quantities of data into formatted files followed by a simultaneous fitting of many uncertain parameters. Because of these complications, studies are often limited to calibration periods of only a few years even when analyzing long time-scale issues such as climate and land-use changes that are difficult or impossible to reproduce. High-level, open source programming languages provide an environment for automating the extraction and processing of various sources of data in addition to the calibration process required by HSPF. Recently developed software can be used to build HSPF input files, run simulations, and postprocess simulation results using the Python programming language. The combination of tools in Python, public data sets on the Internet, and software extensions enables rapid development of long-term, reproducible, and sophisticated models for new hydrologic insight. Herein, the utility of these tools is illustrated by developing an automated 30-year HSPF model with a Nash-Sutcliffe efficiency of 0.88 for monthly flows and 0.75 for daily flows in a simulationmore » time of approximately 3h. The integration of HSPF with a high-level programming language creates opportunities to more rigorously explore model assumptions, calibration methods, and alternative data sets.« less

Authors:
;
Publication Date:
Research Org.:
Argonne National Lab. (ANL), Argonne, IL (United States)
Sponsoring Org.:
USDOE Office of Energy Efficiency and Renewable Energy (EERE) - Bioenergy Technologies Office (BETO)
OSTI Identifier:
1491074
DOE Contract Number:  
AC02-06CH11357
Resource Type:
Journal Article
Journal Name:
Journal of Computing in Civil Engineering
Additional Journal Information:
Journal Volume: 32; Journal Issue: 4; Journal ID: ISSN 0887-3801
Publisher:
American Society of Civil Engineers (ASCE)
Country of Publication:
United States
Language:
English

Citation Formats

Lampert, Dave J, and Wu, May. Automated Approach for Construction of Long-Term, Data-Intensive Watershed Models. United States: N. p., 2018. Web. doi:10.1061/(ASCE)CP.1943-5487.0000762.
Lampert, Dave J, & Wu, May. Automated Approach for Construction of Long-Term, Data-Intensive Watershed Models. United States. doi:10.1061/(ASCE)CP.1943-5487.0000762.
Lampert, Dave J, and Wu, May. Sun . "Automated Approach for Construction of Long-Term, Data-Intensive Watershed Models". United States. doi:10.1061/(ASCE)CP.1943-5487.0000762.
@article{osti_1491074,
title = {Automated Approach for Construction of Long-Term, Data-Intensive Watershed Models},
author = {Lampert, Dave J and Wu, May},
abstractNote = {Watershed models such as the Hydrological Simulation Program in FORTRAN (HSPF) are frequently used to analyze large-scale water quantity and water quality issues. The construction of HSPF models is a difficult and time-consuming process, because it requires compilation of extensive quantities of data into formatted files followed by a simultaneous fitting of many uncertain parameters. Because of these complications, studies are often limited to calibration periods of only a few years even when analyzing long time-scale issues such as climate and land-use changes that are difficult or impossible to reproduce. High-level, open source programming languages provide an environment for automating the extraction and processing of various sources of data in addition to the calibration process required by HSPF. Recently developed software can be used to build HSPF input files, run simulations, and postprocess simulation results using the Python programming language. The combination of tools in Python, public data sets on the Internet, and software extensions enables rapid development of long-term, reproducible, and sophisticated models for new hydrologic insight. Herein, the utility of these tools is illustrated by developing an automated 30-year HSPF model with a Nash-Sutcliffe efficiency of 0.88 for monthly flows and 0.75 for daily flows in a simulation time of approximately 3h. The integration of HSPF with a high-level programming language creates opportunities to more rigorously explore model assumptions, calibration methods, and alternative data sets.},
doi = {10.1061/(ASCE)CP.1943-5487.0000762},
journal = {Journal of Computing in Civil Engineering},
issn = {0887-3801},
number = 4,
volume = 32,
place = {United States},
year = {2018},
month = {7}
}