Quantile estimation for time series with applications to the Swiss precipitation records (supported by the SNF - Division of Mathematics, Natural & Engineering Sciences)
Due to the adverse impacts of extreme weather phenomena in practically all spheres of
life, there is an increasing interest in modeling long term stochastic variations in
climate events. As regards heavy precipitation,
recent studies indicate that the frequencies of such events may have
increased in the Alpine region requiring a proper understanding of such phenomena
through estimation and prediction of the probability distributions
of precipitation events. WSL is involved in a joint collaboration with the Institute of
Climate Research (ETHZ) and the Department of Mathematics (EPFL)
to develop statistical methods to assess the changing precipitation patterns in Switzerland. Collaborators:
Modeling assimilation activity
using Chlorophyll a fluorescence: Diagnosis & Prediction
The aim of this project is to model typical behavior as opposed to behavior under stress of plants via analysis of
chlorophyll a fluorescence curves. Responses of different plant species under varying treatment conditions will be examined in the light of their photosynthetic activities.
The statistical methods to be used here include
time series analysis of repeated measures data. In paricular, fluorescence
measurements on leaves will be taken for two seconds giving rise to some
1200 observations. Typically, the fluorescence
curves exhibit a fast rise to a maximum followed by a slow decline to a
steady state. One of the specific aims here will be to estimate the theoretical
(deterministic trend) fluorescence curve as a function of time and examine how its
shape may change depending on the gradual effect of heavy metals and acid rains.
Collaborators:
Tree ring web and alternative chronologies
Construction of alternatives chronologies involves development of appropriate
statistical models that enable assessment of variabilities as well as
quantiles (in particular extreme quantiles) of the various synchronized
tree ring parameters. Selected alternative chronologies will be made available on the Tree Ring Web in addition to the standard (mean) chronologies. This research has among others, implications
for research in climate change. Collaborators:
Statistics of tree rings - Theory, Software, Web
New time series approaches will be used to assess the long-term behavior of the
various tree ring parameters. Special attention will be given to (a) estimation
of dominating frequencies ("is there a pulse?"), (b) alpha-quantile chronologies (the applied
aspects for this topic will be handled in the previous project, whereas the
theoretical model building will be done within this project),
(c) the sample size issue revisited ("how much more do we learn by taking
additional samples?"),
(d) correlations with climate change (partial linear models and related
approaches). Collaborators:
Time series methods - applications in ecology, finance, etc.
Building time series models that would distinguish between deterministic
as opposed to spurious (due to correlations) trends is important for many
environmental problems. Another concern is to merge physical models with
empirical evidences so as to bridge gaps between theory and data.
Some of the fields of applications for this researh include environment (studies of the polar
ice thickness, global warming, local and global climate variations,
plant physiology), finance (stock market assessments) and health.
Collaborators: