Transcript Montse Fuentes

Research directions in
climate change
Montserrat Fuentes
Statistics Department NCSU
SAMSI workshop,
September 14, 2009.
Observed data and a climate
simulation
Graph provided by D. Nychka
Calibrating general circulation
climate models.
Calibration of climate models is generally
limited to matching the mean and variance
of models and data.
Need to consider different quantiles of
models and data to compare frequency of
extreme events
Spatial quantile
regression might provide a more
comprehensive evaluation.
Calibration plots for temperature and wind speed in the South-East.
Using spatial quantile regression (Reich, Fuentes, Dunson , 2009).
Graph provided by Nychka
Estimating future climate:
Ensemble climate prediction.
Multi model ensembles help with statistical
assessment of model performance. But,
how do we weight different circulation
models?
Bayesian model averaging does not have
physical justification, can they provide
helpful information about model
performace?
Graph provided by Nychka
Impact of climate change on
weather extremes.
 Most of the emphasis has been on global
mean temperatures. Not much is known
about the impact of climate change on
regional weather extremes.
 Modeling of extreme events across space is
challenging
Need of new methods for
spatial modeling of extreme events.
IPCC (2001)
The IPCC (2001) has estimated that the global average
temperature will rise by several degrees centigrade during this
century. There is unavoidable uncertainty in this estimate.
There is a lot of emphasis on global averages. What about
regional extreme temperatures?
Fuentes, Henry and Reich (2009).
Spatial modeling of maximum temperatures
Going beyond
Gaussian models.
Spatial modeling of extremes: important and active area of
research.
Impact of climate change on air
pollution regulation.
Due to the strong dependence on weather
conditions. Ozone levels may be sensitive to
climate change.
Using future numerical climate models, we
could forecast potential future increases or
decreases in ozone levels.
Statistical models for forecasting air pollution.
Reich, Fuentes and Dunson (2009)
Probabilities that the 3 year (2003-2005 left graph; and 20412043 right) average of the fourth highest daily max. 8-hour
average ozone exceeds 75 ppb.
Reich, Fuentes and Dunson
(2009). Bayesian spatial quantile
regression.
Impact of climate change on
human health.
There is great interest in studying the potential effect
of climate change on ozone levels, and how this
change may affect public health.
There is also public health consequences of rising
sea-levels.
Climate change could also alter the geographic
range (latitude and altitude) and seasonality of
certain infectious diseases.
Spatial environmental health epidemiological
modeling. Space-time dynamic risk parameters.
Topics
 Calibrating general circulation climate models. Spatial
quantile regression.
 Estimating future climate: Ensemble climate prediction.
How to weight different circulation models?
 Impact of climate change on weather extremes. Spatial
modelling of extreme events.
 Impact of climate change on air pollution regulation.
Statistical models for forecasting pollution.
 Impact of climate change on human health. Spatial
environmental health modeling.