Developing regression models for predicting pan evaporation from climatic data - A comparison of multiple least-squares, principal components, and partial least-squares approaches

Gicy M. Kovoor, Lakshman Nandagiri

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Regression models for predicting daily pan evaporation depths from climatic data were developed using three multivariate approaches: multiple least-squares regression (MLR), principal components regression (PCR), and partial least-squares (PLS) regression. The objective was to compare the prediction accuracies of regression models developed by these three approaches using historical climatic datasets of four Indian sites that are located in distinctly different climatic regimes. In all cases (three approaches applied to four climatic datasets), regression models were developed using a part of the data and subsequently validated with the remaining data. Results indicated that although performances of the regression models varied from one climate to another, more or less similar prediction accuracies were obtained by all three approaches, and it was difficult to identify the best approach based on performance statistics. However, the final forms of the regression models developed by the three approaches differed substantially from one another. In all cases, the models derived using PLS contained the smallest number of predictor variables; between two to three out of a possible maximum of six predictor variables. The MLR approach yielded models with three to six predictor variables, and PCR models included all six predictor variables. This implies that the PLS regression models are the most parsimonious in terms of input data required for estimating e pan from climate variables, and yet yield predictions that are almost as accurate as the more data-intensive MLR and PCR models.

Original languageEnglish
Pages (from-to)444-454
Number of pages11
JournalJournal of Irrigation and Drainage Engineering
Volume133
Issue number5
DOIs
Publication statusPublished - 01-10-2007

Fingerprint

Least-Squares Analysis
evaporation
least squares
Evaporation
Climate
prediction
comparison
climate
statistics
Statistics

All Science Journal Classification (ASJC) codes

  • Agricultural and Biological Sciences (miscellaneous)
  • Water Science and Technology
  • Civil and Structural Engineering

Cite this

@article{f24bf128cdce442187c4a62fd0061cb0,
title = "Developing regression models for predicting pan evaporation from climatic data - A comparison of multiple least-squares, principal components, and partial least-squares approaches",
abstract = "Regression models for predicting daily pan evaporation depths from climatic data were developed using three multivariate approaches: multiple least-squares regression (MLR), principal components regression (PCR), and partial least-squares (PLS) regression. The objective was to compare the prediction accuracies of regression models developed by these three approaches using historical climatic datasets of four Indian sites that are located in distinctly different climatic regimes. In all cases (three approaches applied to four climatic datasets), regression models were developed using a part of the data and subsequently validated with the remaining data. Results indicated that although performances of the regression models varied from one climate to another, more or less similar prediction accuracies were obtained by all three approaches, and it was difficult to identify the best approach based on performance statistics. However, the final forms of the regression models developed by the three approaches differed substantially from one another. In all cases, the models derived using PLS contained the smallest number of predictor variables; between two to three out of a possible maximum of six predictor variables. The MLR approach yielded models with three to six predictor variables, and PCR models included all six predictor variables. This implies that the PLS regression models are the most parsimonious in terms of input data required for estimating e pan from climate variables, and yet yield predictions that are almost as accurate as the more data-intensive MLR and PCR models.",
author = "Kovoor, {Gicy M.} and Lakshman Nandagiri",
year = "2007",
month = "10",
day = "1",
doi = "10.1061/(ASCE)0733-9437(2007)133:5(444)",
language = "English",
volume = "133",
pages = "444--454",
journal = "Journal of Irrigation and Drainage Engineering - ASCE",
issn = "0733-9437",
publisher = "American Society of Civil Engineers (ASCE)",
number = "5",

}

TY - JOUR

T1 - Developing regression models for predicting pan evaporation from climatic data - A comparison of multiple least-squares, principal components, and partial least-squares approaches

AU - Kovoor, Gicy M.

AU - Nandagiri, Lakshman

PY - 2007/10/1

Y1 - 2007/10/1

N2 - Regression models for predicting daily pan evaporation depths from climatic data were developed using three multivariate approaches: multiple least-squares regression (MLR), principal components regression (PCR), and partial least-squares (PLS) regression. The objective was to compare the prediction accuracies of regression models developed by these three approaches using historical climatic datasets of four Indian sites that are located in distinctly different climatic regimes. In all cases (three approaches applied to four climatic datasets), regression models were developed using a part of the data and subsequently validated with the remaining data. Results indicated that although performances of the regression models varied from one climate to another, more or less similar prediction accuracies were obtained by all three approaches, and it was difficult to identify the best approach based on performance statistics. However, the final forms of the regression models developed by the three approaches differed substantially from one another. In all cases, the models derived using PLS contained the smallest number of predictor variables; between two to three out of a possible maximum of six predictor variables. The MLR approach yielded models with three to six predictor variables, and PCR models included all six predictor variables. This implies that the PLS regression models are the most parsimonious in terms of input data required for estimating e pan from climate variables, and yet yield predictions that are almost as accurate as the more data-intensive MLR and PCR models.

AB - Regression models for predicting daily pan evaporation depths from climatic data were developed using three multivariate approaches: multiple least-squares regression (MLR), principal components regression (PCR), and partial least-squares (PLS) regression. The objective was to compare the prediction accuracies of regression models developed by these three approaches using historical climatic datasets of four Indian sites that are located in distinctly different climatic regimes. In all cases (three approaches applied to four climatic datasets), regression models were developed using a part of the data and subsequently validated with the remaining data. Results indicated that although performances of the regression models varied from one climate to another, more or less similar prediction accuracies were obtained by all three approaches, and it was difficult to identify the best approach based on performance statistics. However, the final forms of the regression models developed by the three approaches differed substantially from one another. In all cases, the models derived using PLS contained the smallest number of predictor variables; between two to three out of a possible maximum of six predictor variables. The MLR approach yielded models with three to six predictor variables, and PCR models included all six predictor variables. This implies that the PLS regression models are the most parsimonious in terms of input data required for estimating e pan from climate variables, and yet yield predictions that are almost as accurate as the more data-intensive MLR and PCR models.

UR - http://www.scopus.com/inward/record.url?scp=39649120346&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=39649120346&partnerID=8YFLogxK

U2 - 10.1061/(ASCE)0733-9437(2007)133:5(444)

DO - 10.1061/(ASCE)0733-9437(2007)133:5(444)

M3 - Article

VL - 133

SP - 444

EP - 454

JO - Journal of Irrigation and Drainage Engineering - ASCE

JF - Journal of Irrigation and Drainage Engineering - ASCE

SN - 0733-9437

IS - 5

ER -