An open software environment for hydrological model assessment and development

被引:117
作者
Andrews, F. T. [1 ,3 ]
Croke, B. F. W. [1 ,2 ,3 ]
Jakeman, A. J. [1 ,2 ]
机构
[1] Australian Natl Univ, Integrated Catchment Assessment & Management iCAM, Fenner Sch Environm & Soc, Canberra, ACT 0200, Australia
[2] Flinders Univ S Australia, Natl Ctr Groundwater Res & Training, Sch Environm, Adelaide, SA 5001, Australia
[3] Australian Natl Univ, Dept Math, Canberra, ACT 0200, Australia
关键词
Model evaluation; Hydrological models; Modelling frameworks; Unit hydrograph; Event separation; R; STANDARDIZED ASSESSMENT; UNIT-HYDROGRAPH; TOP-DOWN; UNCERTAINTY; PERFORMANCE; STREAMFLOW; DYNAMICS;
D O I
10.1016/j.envsoft.2011.04.006
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
The hydromad (Hydrological Model Assessment and Development) package provides a set of functions which work together to construct, manipulate, analyse and compare hydrological models. The class of hydrological models considered are dynamic, spatially-aggregated conceptual or statistical models. The package functions are designed to fit seamlessly into the R system, and builds on its powerful data manipulation and analysis capabilities. The framework used in the package encourages a separation of model components based on Unit Hydrograph theory; many published models are consistent with this and implementations of several are included. For comparative assessment, model performance can be analysed over time and with respect to covariates to reveal systematic biases. Support has been built in for event-based analysis of data and assessment of model performance. Fit statistics can be defined by choices of (1) temporal scale and aggregation function; (2) weighting and transformation; and (3) reference model. One can define new Soil Moisture Accounting models, routing models, calibration methods, objective functions, and evaluation statistics, while retaining as much of the default framework as is useful. And as the package code is available under a free software licence, one always has the freedom to adapt it as required. Use of the software is demonstrated in a case study of the Queanbeyan River catchment in South-East Australia. Crown Copyright (C) 2011 Published by Elsevier Ltd. All rights reserved.
引用
收藏
页码:1171 / 1185
页数:15
相关论文
共 65 条
  • [1] ANDREWS FT, 2010, P IEMSS 5 BIENN M SU
  • [2] Ardia D., 2010, DEOPTIM PACKAGE DIFF
  • [3] A top-down framework for watershed model evaluation and selection under uncertainty
    Bai, Yaoling
    Wagener, Thorsten
    Reed, Patrick
    [J]. ENVIRONMENTAL MODELLING & SOFTWARE, 2009, 24 (08) : 901 - 916
  • [4] Bates D, 2008, TECHNOMETRICS, V50, P439, DOI 10.1198/004017008000000488
  • [5] Becker R.A., 1988, NEW S LANGUAGE PROGR
  • [6] A manifesto for the equifinality thesis
    Beven, K
    [J]. JOURNAL OF HYDROLOGY, 2006, 320 (1-2) : 18 - 36
  • [7] On doing better hydrological science
    Beven, Keith
    [J]. HYDROLOGICAL PROCESSES, 2008, 22 (17) : 3549 - 3553
  • [8] Generalized likelihood uncertainty estimation (GLUE) using adaptive Markov chain Monte Carlo sampling
    Blasone, Roberta-Serena
    Vrugt, Jasper A.
    Madsen, Henrik
    Rosbjerg, Dan
    Robinson, Bruce A.
    Zyvoloski, George A.
    [J]. ADVANCES IN WATER RESOURCES, 2008, 31 (04) : 630 - 648
  • [9] The Australian water balance model
    Boughton, W
    [J]. ENVIRONMENTAL MODELLING & SOFTWARE, 2004, 19 (10) : 943 - 956
  • [10] AN ANALYSIS OF TRANSFORMATIONS
    BOX, GEP
    COX, DR
    [J]. JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY, 1964, 26 (02) : 211 - 252