Logit-normal mixed model for Indian monsoon precipitation

L. R. Dietz, S. Chatterjee

Research output: Contribution to journalArticle

Abstract

Describing the nature and variability of Indian monsoon precipitation is a topic of much debate in the current literature. We suggest the use of a generalized linear mixed model (GLMM), specifically, the logit-normal mixed model, to describe the underlying structure of this complex climatic event. Four GLMM algorithms are described and simulations are performed to vet these algorithms before applying them to the Indian precipitation data. The logit-normal model was applied to light, moderate, and extreme rainfall. Findings indicated that physical constructs were preserved by the models, and random effects were significant in many cases. We also found GLMM estimation methods were sensitive to tuning parameters and assumptions and therefore, recommend use of multiple methods in applications. This work provides a novel use of GLMM and promotes its addition to the gamut of tools for analysis in studying climate phenomena.

Original languageEnglish (US)
Pages (from-to)939-953
Number of pages15
JournalNonlinear Processes in Geophysics
Volume21
Issue number5
DOIs
StatePublished - Sep 12 2014

Fingerprint

monsoons
monsoon
estimation method
climate
Rain
Tuning
tuning
rainfall
simulation

Cite this

Logit-normal mixed model for Indian monsoon precipitation. / Dietz, L. R.; Chatterjee, S.

In: Nonlinear Processes in Geophysics, Vol. 21, No. 5, 12.09.2014, p. 939-953.

Research output: Contribution to journalArticle

@article{e48f18bf4ec6426393d46ecab0227a49,
title = "Logit-normal mixed model for Indian monsoon precipitation",
abstract = "Describing the nature and variability of Indian monsoon precipitation is a topic of much debate in the current literature. We suggest the use of a generalized linear mixed model (GLMM), specifically, the logit-normal mixed model, to describe the underlying structure of this complex climatic event. Four GLMM algorithms are described and simulations are performed to vet these algorithms before applying them to the Indian precipitation data. The logit-normal model was applied to light, moderate, and extreme rainfall. Findings indicated that physical constructs were preserved by the models, and random effects were significant in many cases. We also found GLMM estimation methods were sensitive to tuning parameters and assumptions and therefore, recommend use of multiple methods in applications. This work provides a novel use of GLMM and promotes its addition to the gamut of tools for analysis in studying climate phenomena.",
author = "Dietz, {L. R.} and S. Chatterjee",
year = "2014",
month = "9",
day = "12",
doi = "10.5194/npg-21-939-2014",
language = "English (US)",
volume = "21",
pages = "939--953",
journal = "Nonlinear Processes in Geophysics",
issn = "1023-5809",
publisher = "European Geosciences Union",
number = "5",

}

TY - JOUR

T1 - Logit-normal mixed model for Indian monsoon precipitation

AU - Dietz, L. R.

AU - Chatterjee, S.

PY - 2014/9/12

Y1 - 2014/9/12

N2 - Describing the nature and variability of Indian monsoon precipitation is a topic of much debate in the current literature. We suggest the use of a generalized linear mixed model (GLMM), specifically, the logit-normal mixed model, to describe the underlying structure of this complex climatic event. Four GLMM algorithms are described and simulations are performed to vet these algorithms before applying them to the Indian precipitation data. The logit-normal model was applied to light, moderate, and extreme rainfall. Findings indicated that physical constructs were preserved by the models, and random effects were significant in many cases. We also found GLMM estimation methods were sensitive to tuning parameters and assumptions and therefore, recommend use of multiple methods in applications. This work provides a novel use of GLMM and promotes its addition to the gamut of tools for analysis in studying climate phenomena.

AB - Describing the nature and variability of Indian monsoon precipitation is a topic of much debate in the current literature. We suggest the use of a generalized linear mixed model (GLMM), specifically, the logit-normal mixed model, to describe the underlying structure of this complex climatic event. Four GLMM algorithms are described and simulations are performed to vet these algorithms before applying them to the Indian precipitation data. The logit-normal model was applied to light, moderate, and extreme rainfall. Findings indicated that physical constructs were preserved by the models, and random effects were significant in many cases. We also found GLMM estimation methods were sensitive to tuning parameters and assumptions and therefore, recommend use of multiple methods in applications. This work provides a novel use of GLMM and promotes its addition to the gamut of tools for analysis in studying climate phenomena.

UR - http://www.scopus.com/inward/record.url?scp=84907910069&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84907910069&partnerID=8YFLogxK

U2 - 10.5194/npg-21-939-2014

DO - 10.5194/npg-21-939-2014

M3 - Article

AN - SCOPUS:84907910069

VL - 21

SP - 939

EP - 953

JO - Nonlinear Processes in Geophysics

JF - Nonlinear Processes in Geophysics

SN - 1023-5809

IS - 5

ER -