Determining the importance of model calibration for forecasting absolute/relative changes in streamflow from LULC and climate changes

Rewati Niraula, Thomas Meixner, Laura M. Norman

Research output: Contribution to journalArticle

39 Citations (Scopus)

Abstract

Land use/land cover (LULC) and climate changes are important drivers of change in streamflow. Assessing the impact of LULC and climate changes on streamflow is typically done with a calibrated and validated watershed model. However, there is a debate on the degree of calibration required. The objective of this study was to quantify the variation in estimated relative and absolute changes in streamflow associated with LULC and climate changes with different calibration approaches. The Soil and Water Assessment Tool (SWAT) was applied in an uncalibrated (UC), single outlet calibrated (OC), and spatially-calibrated (SC) mode to compare the relative and absolute changes in streamflow at 14 gaging stations within the Santa Cruz River Watershed in southern Arizona, USA. For this purpose, the effect of 3 LULC, 3 precipitation (P), and 3 temperature (T) scenarios were tested individually. For the validation period, Percent Bias (PBIAS) values were >100% with the UC model for all gages, the values were between 0% and 100% with the OC model and within 20% with the SC model. Changes in streamflow predicted with the UC and OC models were compared with those of the SC model. This approach implicitly assumes that the SC model is "ideal". Results indicated that the magnitude of both absolute and relative changes in streamflow due to LULC predicted with the UC and OC results were different than those of the SC model. The magnitude of absolute changes predicted with the UC and SC models due to climate change (both P and T) were also significantly different, but were not different for OC and SC models. Results clearly indicated that relative changes due to climate change predicted with the UC and OC were not significantly different than that predicted with the SC models. This result suggests that it is important to calibrate the model spatially to analyze the effect of LULC change but not as important for analyzing the relative change in streamflow due to climate change. This study also indicated that model calibration in not necessary to determine the direction of change in streamflow due to LULC and climate change.

Original languageEnglish (US)
Pages (from-to)439-451
Number of pages13
JournalJournal of Hydrology
Volume522
DOIs
StatePublished - Mar 1 2015

Fingerprint

streamflow
land cover
calibration
land use
climate change
watershed
gauge

Keywords

  • Calibration
  • Climate change
  • LULC change
  • Santa cruz watershed
  • Streamflow
  • SWAT

ASJC Scopus subject areas

  • Water Science and Technology

Cite this

Determining the importance of model calibration for forecasting absolute/relative changes in streamflow from LULC and climate changes. / Niraula, Rewati; Meixner, Thomas; Norman, Laura M.

In: Journal of Hydrology, Vol. 522, 01.03.2015, p. 439-451.

Research output: Contribution to journalArticle

@article{9f73e13180a240a6babb23cdb1feae5d,
title = "Determining the importance of model calibration for forecasting absolute/relative changes in streamflow from LULC and climate changes",
abstract = "Land use/land cover (LULC) and climate changes are important drivers of change in streamflow. Assessing the impact of LULC and climate changes on streamflow is typically done with a calibrated and validated watershed model. However, there is a debate on the degree of calibration required. The objective of this study was to quantify the variation in estimated relative and absolute changes in streamflow associated with LULC and climate changes with different calibration approaches. The Soil and Water Assessment Tool (SWAT) was applied in an uncalibrated (UC), single outlet calibrated (OC), and spatially-calibrated (SC) mode to compare the relative and absolute changes in streamflow at 14 gaging stations within the Santa Cruz River Watershed in southern Arizona, USA. For this purpose, the effect of 3 LULC, 3 precipitation (P), and 3 temperature (T) scenarios were tested individually. For the validation period, Percent Bias (PBIAS) values were >100{\%} with the UC model for all gages, the values were between 0{\%} and 100{\%} with the OC model and within 20{\%} with the SC model. Changes in streamflow predicted with the UC and OC models were compared with those of the SC model. This approach implicitly assumes that the SC model is {"}ideal{"}. Results indicated that the magnitude of both absolute and relative changes in streamflow due to LULC predicted with the UC and OC results were different than those of the SC model. The magnitude of absolute changes predicted with the UC and SC models due to climate change (both P and T) were also significantly different, but were not different for OC and SC models. Results clearly indicated that relative changes due to climate change predicted with the UC and OC were not significantly different than that predicted with the SC models. This result suggests that it is important to calibrate the model spatially to analyze the effect of LULC change but not as important for analyzing the relative change in streamflow due to climate change. This study also indicated that model calibration in not necessary to determine the direction of change in streamflow due to LULC and climate change.",
keywords = "Calibration, Climate change, LULC change, Santa cruz watershed, Streamflow, SWAT",
author = "Rewati Niraula and Thomas Meixner and Norman, {Laura M.}",
year = "2015",
month = "3",
day = "1",
doi = "10.1016/j.jhydrol.2015.01.007",
language = "English (US)",
volume = "522",
pages = "439--451",
journal = "Journal of Hydrology",
issn = "0022-1694",
publisher = "Elsevier",

}

TY - JOUR

T1 - Determining the importance of model calibration for forecasting absolute/relative changes in streamflow from LULC and climate changes

AU - Niraula, Rewati

AU - Meixner, Thomas

AU - Norman, Laura M.

PY - 2015/3/1

Y1 - 2015/3/1

N2 - Land use/land cover (LULC) and climate changes are important drivers of change in streamflow. Assessing the impact of LULC and climate changes on streamflow is typically done with a calibrated and validated watershed model. However, there is a debate on the degree of calibration required. The objective of this study was to quantify the variation in estimated relative and absolute changes in streamflow associated with LULC and climate changes with different calibration approaches. The Soil and Water Assessment Tool (SWAT) was applied in an uncalibrated (UC), single outlet calibrated (OC), and spatially-calibrated (SC) mode to compare the relative and absolute changes in streamflow at 14 gaging stations within the Santa Cruz River Watershed in southern Arizona, USA. For this purpose, the effect of 3 LULC, 3 precipitation (P), and 3 temperature (T) scenarios were tested individually. For the validation period, Percent Bias (PBIAS) values were >100% with the UC model for all gages, the values were between 0% and 100% with the OC model and within 20% with the SC model. Changes in streamflow predicted with the UC and OC models were compared with those of the SC model. This approach implicitly assumes that the SC model is "ideal". Results indicated that the magnitude of both absolute and relative changes in streamflow due to LULC predicted with the UC and OC results were different than those of the SC model. The magnitude of absolute changes predicted with the UC and SC models due to climate change (both P and T) were also significantly different, but were not different for OC and SC models. Results clearly indicated that relative changes due to climate change predicted with the UC and OC were not significantly different than that predicted with the SC models. This result suggests that it is important to calibrate the model spatially to analyze the effect of LULC change but not as important for analyzing the relative change in streamflow due to climate change. This study also indicated that model calibration in not necessary to determine the direction of change in streamflow due to LULC and climate change.

AB - Land use/land cover (LULC) and climate changes are important drivers of change in streamflow. Assessing the impact of LULC and climate changes on streamflow is typically done with a calibrated and validated watershed model. However, there is a debate on the degree of calibration required. The objective of this study was to quantify the variation in estimated relative and absolute changes in streamflow associated with LULC and climate changes with different calibration approaches. The Soil and Water Assessment Tool (SWAT) was applied in an uncalibrated (UC), single outlet calibrated (OC), and spatially-calibrated (SC) mode to compare the relative and absolute changes in streamflow at 14 gaging stations within the Santa Cruz River Watershed in southern Arizona, USA. For this purpose, the effect of 3 LULC, 3 precipitation (P), and 3 temperature (T) scenarios were tested individually. For the validation period, Percent Bias (PBIAS) values were >100% with the UC model for all gages, the values were between 0% and 100% with the OC model and within 20% with the SC model. Changes in streamflow predicted with the UC and OC models were compared with those of the SC model. This approach implicitly assumes that the SC model is "ideal". Results indicated that the magnitude of both absolute and relative changes in streamflow due to LULC predicted with the UC and OC results were different than those of the SC model. The magnitude of absolute changes predicted with the UC and SC models due to climate change (both P and T) were also significantly different, but were not different for OC and SC models. Results clearly indicated that relative changes due to climate change predicted with the UC and OC were not significantly different than that predicted with the SC models. This result suggests that it is important to calibrate the model spatially to analyze the effect of LULC change but not as important for analyzing the relative change in streamflow due to climate change. This study also indicated that model calibration in not necessary to determine the direction of change in streamflow due to LULC and climate change.

KW - Calibration

KW - Climate change

KW - LULC change

KW - Santa cruz watershed

KW - Streamflow

KW - SWAT

UR - http://www.scopus.com/inward/record.url?scp=84921482739&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84921482739&partnerID=8YFLogxK

U2 - 10.1016/j.jhydrol.2015.01.007

DO - 10.1016/j.jhydrol.2015.01.007

M3 - Article

AN - SCOPUS:84921482739

VL - 522

SP - 439

EP - 451

JO - Journal of Hydrology

JF - Journal of Hydrology

SN - 0022-1694

ER -