Climate inference on daily rainfall across the Australian continent, 1876-2015

Michael Bertolacci, Edward Cripps, Ori Rosen, John W. Lau, Sally Cripps

Research output: Contribution to journalArticle

Abstract

Daily precipitation has an enormous impact on human activity, and the study of how it varies over time and space, and what global indicators influence it, is of paramount importance to Australian agriculture. We analyze over 294 million daily rainfall measurements since 1876, spanning 17,606 sites across continental Australia. The data are not only large but also complex, and the topic would benefit from a common and publicly available statistical framework. We propose a Bayesian hierarchical mixture model that accommodates mixed discrete-continuous data. The observational level describes site-specific temporal and climatic variation via a mixture-of-experts model. At the next level of the hierarchy, spatial variability of the mixture weights' parameters is modeled by a spatial Gaussian process prior. A parallel and distributed Markov chain Monte Carlo sampler is developed which scales the model to large data sets. We present examples of posterior inference on the mixture weights, monthly intensity levels, daily temporal dependence, offsite prediction of the effects of climate drivers and long-term rainfall trends across the entire continent. Computer code implementing the methods proposed in this paper is available as an R package.

Original languageEnglish
Pages (from-to)683-712
Number of pages30
JournalAnnals of Applied Statistics
Volume13
Issue number2
DOIs
Publication statusPublished - Jun 2019

Cite this

Bertolacci, Michael ; Cripps, Edward ; Rosen, Ori ; Lau, John W. ; Cripps, Sally. / Climate inference on daily rainfall across the Australian continent, 1876-2015. In: Annals of Applied Statistics. 2019 ; Vol. 13, No. 2. pp. 683-712.
@article{288d60bd24c2436aa54d8de02daeb1a1,
title = "Climate inference on daily rainfall across the Australian continent, 1876-2015",
abstract = "Daily precipitation has an enormous impact on human activity, and the study of how it varies over time and space, and what global indicators influence it, is of paramount importance to Australian agriculture. We analyze over 294 million daily rainfall measurements since 1876, spanning 17,606 sites across continental Australia. The data are not only large but also complex, and the topic would benefit from a common and publicly available statistical framework. We propose a Bayesian hierarchical mixture model that accommodates mixed discrete-continuous data. The observational level describes site-specific temporal and climatic variation via a mixture-of-experts model. At the next level of the hierarchy, spatial variability of the mixture weights' parameters is modeled by a spatial Gaussian process prior. A parallel and distributed Markov chain Monte Carlo sampler is developed which scales the model to large data sets. We present examples of posterior inference on the mixture weights, monthly intensity levels, daily temporal dependence, offsite prediction of the effects of climate drivers and long-term rainfall trends across the entire continent. Computer code implementing the methods proposed in this paper is available as an R package.",
keywords = "Climate, rainfall, Australia, mixture-of-experts, Gaussian processes, parallel and distributed computing, LAND-COVER CHANGE, DAILY PRECIPITATION, WESTERN-AUSTRALIA, MODEL, TEMPERATURE, VARIABILITY, ASYMMETRY, MIXTURES, IMPACT, SOUTH",
author = "Michael Bertolacci and Edward Cripps and Ori Rosen and Lau, {John W.} and Sally Cripps",
year = "2019",
month = "6",
doi = "10.1214/18-AOAS1218",
language = "English",
volume = "13",
pages = "683--712",
journal = "Annals of Applied Statistics",
issn = "1932-6157",
publisher = "Institute of Mathematical Statistics",
number = "2",

}

Climate inference on daily rainfall across the Australian continent, 1876-2015. / Bertolacci, Michael; Cripps, Edward; Rosen, Ori; Lau, John W.; Cripps, Sally.

In: Annals of Applied Statistics, Vol. 13, No. 2, 06.2019, p. 683-712.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Climate inference on daily rainfall across the Australian continent, 1876-2015

AU - Bertolacci, Michael

AU - Cripps, Edward

AU - Rosen, Ori

AU - Lau, John W.

AU - Cripps, Sally

PY - 2019/6

Y1 - 2019/6

N2 - Daily precipitation has an enormous impact on human activity, and the study of how it varies over time and space, and what global indicators influence it, is of paramount importance to Australian agriculture. We analyze over 294 million daily rainfall measurements since 1876, spanning 17,606 sites across continental Australia. The data are not only large but also complex, and the topic would benefit from a common and publicly available statistical framework. We propose a Bayesian hierarchical mixture model that accommodates mixed discrete-continuous data. The observational level describes site-specific temporal and climatic variation via a mixture-of-experts model. At the next level of the hierarchy, spatial variability of the mixture weights' parameters is modeled by a spatial Gaussian process prior. A parallel and distributed Markov chain Monte Carlo sampler is developed which scales the model to large data sets. We present examples of posterior inference on the mixture weights, monthly intensity levels, daily temporal dependence, offsite prediction of the effects of climate drivers and long-term rainfall trends across the entire continent. Computer code implementing the methods proposed in this paper is available as an R package.

AB - Daily precipitation has an enormous impact on human activity, and the study of how it varies over time and space, and what global indicators influence it, is of paramount importance to Australian agriculture. We analyze over 294 million daily rainfall measurements since 1876, spanning 17,606 sites across continental Australia. The data are not only large but also complex, and the topic would benefit from a common and publicly available statistical framework. We propose a Bayesian hierarchical mixture model that accommodates mixed discrete-continuous data. The observational level describes site-specific temporal and climatic variation via a mixture-of-experts model. At the next level of the hierarchy, spatial variability of the mixture weights' parameters is modeled by a spatial Gaussian process prior. A parallel and distributed Markov chain Monte Carlo sampler is developed which scales the model to large data sets. We present examples of posterior inference on the mixture weights, monthly intensity levels, daily temporal dependence, offsite prediction of the effects of climate drivers and long-term rainfall trends across the entire continent. Computer code implementing the methods proposed in this paper is available as an R package.

KW - Climate

KW - rainfall

KW - Australia

KW - mixture-of-experts

KW - Gaussian processes

KW - parallel and distributed computing

KW - LAND-COVER CHANGE

KW - DAILY PRECIPITATION

KW - WESTERN-AUSTRALIA

KW - MODEL

KW - TEMPERATURE

KW - VARIABILITY

KW - ASYMMETRY

KW - MIXTURES

KW - IMPACT

KW - SOUTH

U2 - 10.1214/18-AOAS1218

DO - 10.1214/18-AOAS1218

M3 - Article

VL - 13

SP - 683

EP - 712

JO - Annals of Applied Statistics

JF - Annals of Applied Statistics

SN - 1932-6157

IS - 2

ER -