Environmental Engineering Reference
In-Depth Information
3.4 A Case Study of 4D-VAR Data Assimilation
in Southern Europe
Julius Vira, Marje Prank, and Mikhail Sofiev
Finnish Meteorological Institute, P.O. Box 503, FI-00101 Helsinki, Finland
Abstract
This paper describes a case study on variational data assimilation of
in-situ measurements at 260 stations in Central and Southern Europe, using the
SILAM chemistry-transport model. The aim of the experiment was to investigate
the emission distribution, and both initial concentration and emission correction
factor were included in estimated parameters. The main improvement in forecasts
due to assimilation was seen at sites with the worst model-measurement agreement.
Throughout the assimilation period, the estimated emission rates for Mt. Etna and
the surrounding volcanic areas were substantially smaller than those derived from
the emission inventory. Subtler emission corrections were obtained for industrialised
areas in Central Europe.
Keywords
Data assimilation, chemistry-transport models, sulphur dioxide, emissions
1. Introduction
Extending the concept and methods of data assimilation into atmospheric chemistry-
transport models has been a topic of active research over the last decades. In addition
to operational forecasting of air quality, data assimilation in air pollution modelling
is closely related to atmospheric inverse problems for source localisation and
apportionment. However, this requires using advanced variational or sequential
assimilation methods.
The evolution of atmospheric pollutants is strongly affected by the distribution
of emission sources, and consequently, estimating the emission sources through
data assimilation has been the focus of recent studies including [1]. A common
approach has been to consider an emission distribution derived from inventories,
and use data assimilation to estimate a correction factor.
This paper describes an assimilation experiment based on in-situ measurements
of SO
2
and the 4D variational method with the emission correction approach. In
particular, we explore the potential of a dataset with limited geographic coverage
in emission estimation and operational forecasting.
