ISRR INTERNATIONAL SYMPOSIUM ON REGIONAL REANALYSIS
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
INTERNATIONAL SYMPOSIUM
ISRR
2018
ON REGIONAL REANALYSIS
17 -19 July 2018, Bonn, Germany
Bonn
Germany
Scientific program
Organized by
Sponsored by
IMPLEMENTED BY
Climate Change
ServiceContents Table of Contents
Oral Present
Contents
Introduction 3
Oral Presentations - Tuesday - 17 July 2018 4
Oral Presentations - Wednesday - 18 July 2018 14
Oral Presentations - Thursday - 19 July 2018 28
Poster Presentations 34
2ISRR
2018
Bonn
Germany
Introduction
R egional reanalysis is a relatively new field of research. Starting with the North Amer-
ican Regional Reanalysis (NARR) and the Arctic System Reanalysis (ASR) just over
ten years ago, the community is now taking a much broader shape with multiple regional
reanalysis projects having been implemented and data sets produced in recent years.
While this new developments were primarily concentrated in Europe, we now see more
and more projects from other parts of the world.
Further, reanalysis users are increasingly recognizing the benefits of regional reanalysis
with its increased resolution in multiple contexts such as an enhanced representation
of weather and climate extremes or as much more suitable input for subsequent mod-
elling, e.g., hydrology, compared to global reanalyses. Recently, the ongoing transition
towards renewable energies lead to a considerable further increase in users of regional
reanalyses.
Therefore, it is our great pleasure to welcome the growing community of developers,
producers and users of regional reanalysis to the first International Symposium on Re-
gional Reanalysis (ISRR). We are happy that we are all able to come together in Bonn
for this three-day international symposium and it is our hope that this conference is a
trigger for a continuing series of conferences or events every couple of years. It is our
aim to enhance the understanding of the potential of regional reanalysis and to foster
the disussion between data providers and users in order to further improve regional
reanalysis especially with respect to its applications.
The Scientific Committee consisting of Eric Bazile (MeteoFrance), Andrea Kaiser-Weiss
(DWD), Frank Kaspar (DWD) and Per Unden (SMHI) chaired by Jan Keller (HErZ) has
structured the symposium around three main topics, namely:
• Regional reanalysis projects on varios scales
• Evaluation of regional reanalysis
• Applications of regional reanalysis
The ISRR is organised by the Hans-Ertel-Centre for Weather Research (HErZ funded by
the German Meteorological Service DWD) and the University of Bonn.
We look forward to meeting you!
The ISRR Organisation and Scientific Committees
We would like to thank all our partners and supporters of the International Symposium
on Regional Reanalysis in Bonn:
IMPLEMENTED BY
Climate Change
Service
3Oral Presentations Tuesday - 17 July 2018
ISRR
2018
Bonn
Germany
TU.0I Development and applications of region-
al reanalyses for Europe and Germany
Oral Presentations
based on DWD’s NWP models: Status and
outlook
Tuesday
Authors: Frank Kaspar1, Michael Borsche1, Andrea K. Kaiser-Weiss1, Jan
D. Keller1,2, Maarit Lockhoff 3, Deborah Niermann1, Roland
Potthast1, Thomas Roesch1,2, Sabrina Wahl2,3
1
Affiliations: Deutscher Wetterdienst, Offenbach, Germany, 2Hans-Er-
tel-Centre for Weather Research, Germany, 3Meteorological
Institute, University of Bonn, Germany
I
Oral Presentations
n recent years, Germany’s national meteorological service (DWD) has start-
ed several activities aiming at the development and evaluation of regional
Wednesday
reanalyses as a tool for climate monitoring. The first step was the devel-
opment of a COSMO-based reanalysis within the Hans-Ertel-Centre for
Weather Research. The COSMO model (Consortium for Small-scale Mod-
eling) is a non-hydrostatic limited-area atmospheric model, that is used for
operational weather forecasting as well as for research applications at the
German Meteorological Service (Deutscher Wetterdienst, DWD) and other
members of the consortium (see http://www.cosmo-model.org/ ). The model
was the center of the German regional reanalysis activities and was also used
to generate a set of regional reanalyses within two major activities: (1) The
HErZ branch on climate monitoring and diagnostics is a DWD-funded research
Oral Presentations
group located at the Universities of Bonn and Cologne. Its aim is to analyze
the potential of regional reanalysis systems as a tool for climate monitoring
Thursday
in comparison to global reanalyses and other data sets. A 6 km deterministic
reanalysis for the European CORDEX-domain and a 2 km reanalysis for
Central Europe has been produced within the Hans-Ertel-Centre for Weath-
er Research (HErZ). Current developments include an ultra high-resolution
local reanalysis system comprising a coupled terrestrial system approach. (2)
Within the EU-funded project UERRA an approach for a COSMO ensemble
reanalysis was developed. An ensemble dataset with 12 km resolution was
produced and is currently compared to regional reanalyses of other Eu-
ropean partners. The application potential of these reanalyses datasets is
Tuesday / Wednesday
Poster Presentations
currently evaluated within several additional projects. All these current re-
analyses are based on nudging as data assimilation scheme. In the meantime,
progress has been made in the NWP models and data assimilation techniques.
The current operational data assimilation system at DWD consists of a global
component, using the ICON model and a combination of LETKF+EnVar for
the ensemble and deterministic part of the analysis, and a regional compo-
nent, combining a LETKF and the kilometer scale COSMO model (KENDA).
Currently scenarios are under investigation how these components can be
5International Symposium on Regional Reanalysis
used for the production of future reanalyses. The presentation will give an
overview over the existing regional reanalyses, results from quality assess-
Oral Presentations
ments, application examples and an outlook on future plans.
Tuesday
TU.02 China Regional Reanalysis : One-year Pre-
liminary Experiments and Evaluation of
First Stage (1998 ~ 2007) Datasets
Authors: Qi Zhang1, Jianping Tang1,Yinong Pan1
1
Affiliation: School of Atmospheric Science, Nanjing University
B y taking the advantage of limited area meteorological numerical model,
Oral Presentations
data assimilation or nudging method, regional reanalysis aims at generating
a series of homogeneous estimation of regional atmospheric variables with
Wednesday
a higher temporal and spatial resolution than global reanalysis. As is the fact
that regional reanalysis dataset has already been generated in many areas
e.g. North America [Mesinger et al., 2006], Arctic [Bromwich et al., 2016],
Europe [Bollmeyer, Keller, et al., 2015; Dahlgren et al., 2016], The demand of
generating a high spatial and temporal resolution regional reanalysis dataset is
prevailing.
China Regional Reanalysis project (CNRR) is a program conducted by the
School of Atmospheric Science, Nanjing University which aims to generate a
series of relatively long time sequence (1998 ~ now), high spatial (18km
horizontal resolution) and temporal (1-hourly) resolution regional reanalysis
Oral Presentations
datasets for China mainland.
Based on results we gathered from preliminary experiments [Zhang et al.,
Thursday
2017], three types of datasets are now under production: CNRR-ASOBS,
dataset assimilating conventional and satellite radiance data via 3Dvar method;
CNRR-SN, dataset using spectral nudging method; CNRR-ASERA, dataset us-
ing a new method to correct every 6-hourly warm start initial field.
As the first stage of all three datasets (1998 ~ 2007) have ready been gen-
erated, we are now conduction a general evaluation for near surface and up-
per air atmospheric variables. It would be a great honor if we are permitted
to share our latest progress about CNRR via a poster and get constructive
Tuesday / Wednesday
Poster Presentations
advices or suggestions through ISRR.
TU.03 Evaluation of Australian regional atmo-
spheric reanalysis
Authors: Chun-Hsu Su1, Nathan Eizenberg1, Peter Steinle1, Doerte Jakob1,
Paul Fox-Hughes2
1
Affiliations: Bureau of Meteorology, Melbourne, Australia, 2Bureau of Mete-
6ISRR
2018
Bonn
Germany
orology, Hobart, Australia
T
Oral Presentations
he Bureau of Meteorology Atmospheric high-resolution Regional Re-
analysis for Australia (BARRA, http://www.bom.gov.au/research/projects/
Tuesday
reanalysis) is the first atmospheric reanalysis produced by the Bureau for
the Australian region. When completed in 2019, BARRA will provide over 25
years of hourly reconstructions of past atmospheric conditions from 1990
onwards. BARRA’s 12 km regional reanalysis (BARRA-R) over the Australia
and New Zealand regions is bounded by the ERA-Interim reanalysis. BARRA
also includes multiple convection permitting (1.5 km) dynamical downscal-
ing analyses (BARRA-xx, where xx labels downscaled domain) over smaller
sub-domains, nested within BARRA-R, to provide additional fine-scale detail.
BARRA is based on fixed deterministic atmospheric (4DVAR) and land sur-
Oral Presentations
face (EKF) assimilation systems used at the UK Met Office (UERRA, Uncer-
tainties in Ensembles of Regional Reanalyses), and model nesting and Unified
Wednesday
Model configurations based on the Bureau’s operational numerical prediction
system (ACCESS, Australian Community Climate and Earth-System Simula-
tor). In contrast to the global reanalyses, high-resolution information on the
land surface (tree heights, topography, etc) was used along with additional ob-
servations from automatic weather stations in Australia and New Zealand,
and locally derived satellite wind vectors over Australia and New Zealand.
BARRA is initialized with the soil temperature and moisture fields derived
from a lower resolution (60 km) offline run of the Joint UK Land Environ-
ment Simulator (JULES) forced by ERA-Interim atmospheric variables.
The finer horizontal resolutions and the use of additional surface observa-
Oral Presentations
tions are shown to add realistic fine-scale detail, at successive stages from
ERA-Interim to BARRA-R and to BARRA-xx. This is most likely associated
Thursday
with a more realistic representation of surface forcing (orography, tree heights,
coastal regions, etc), a better simulation of the atmospheric circulation and
gradients, and more explicit representations of fine scale processes (e.g., sea
breezes, convective systems). We demonstrate the distinct advantage in using
BARRA-xx over ERA-Interim for near surface wind, temperature and pre-
cipitation. In this presentation, we report our skill assessments of BARRA-R
and BARRA-xx data, focussing on forcing variables (e.g. screen temperature
and humidity, 10 m wind, and precipitation) that influence extreme weather
events, and highlight the added value BARRA can provide over coarser global
Tuesday / Wednesday
Poster Presentations
reanalyses and traditional gridded climate datasets.
TU.04 An EOF-VARX model for statistical down-
scaling
Authors: Ci-Ren Jiang1, Lu-Hung Chen2
1
Affiliations: Academia Sinica, Taipei, Taiwan, 2National Chung-Hsing Univer-
7International Symposium on Regional Reanalysis
sity, Taichung, Taiwan
W
Oral Presentations
e suggest a strategy for downscaling daily mean temperature that
incorporates their temporal relationships. The strategy is based on
Tuesday
empirical orthogonal functions (EOFs) and vector autoregressive model with
exogenous variables (VARX). The missing values in observational data are re-
constructed by data interpolating empirical orthogonal functions (DINEOF,
Beckers and Rixen 2003), and the temporal information for daily mean
temperatures are utilized by VARX models. We test the strategy in some
empirical studies for Germany and Taiwan. Results show that our EOF-VARX
model provides more accurate downscaled results in terms of root-mean-
square error and correlation scores.
Oral Presentations
TU.05 A 140-year high-resolution meteorologi-
Wednesday
cal reanalysis over France through offline
data assimilation in an ensemble of down-
scaled reconstructions from 20CR
Authors: Alexandre Devers1, Jean-Philippe Vidal1, Claire Lauvernet1, Ben-
jamin Graff2
1
Affiliations: Irstea, UR RiverLy,Villeurbanne, France, 2Compagnie Nationale
du Rhone (CNR), Lyon, France
T he knowledge of historical French weather has recently been improved
Oral Presentations
through the development of the SCOPE (Spatially COherent Proba-
bilistic Extended) Climate reconstruction, a high-resolution daily ensemble
Thursday
reconstruction of precipitation and temperature covering the period 1871-
2012, which is based on a statistical downscaling of the Twentieth Century
Reanalysis (Caillouet et al., 2016, 2017). Historical surface observations of
precipitation and temperature do exist in France, though rather scarce and
sparse in the beginning of the period considered, but this information does
not currently feed SCOPE Climate reconstructions. The goal of this study
is therefore to assimilate these historical observations into SCOPE Climate
reconstructions in order to build a 140-year meteorological reanalysis over
France.
Tuesday / Wednesday
Poster Presentations
This study considers an offline data assimilation set-up (ie, without running
any model) that have successfully been used in recent paleoclimate stud-
ies, i.e. at much larger temporal and spatial scales (see e.g. Bhend et al., 2012).
The data assimilation method applied here is based on the Ensemble Kalman
Filter. An anisotropic localization procedure based on the SCOPE Climate
background climatology is applied on both temperature and precipitation. A
Gaussian anamorphosis is moreover applied on daily precipitation.
The data assimilation method is here implemented to build FYRE Climate
8ISRR
2018
Bonn
Germany
(French hYdrometeorological Reanalysis), a daily 8-km gridded meteorologi-
cal reanalysis over France for the 1871-2012 period. Results show that: (1)
Oral Presentations
the reanalysis has a much lower uncertainty than the initial SCOPE Climate
reconstructions, (2) the uncertainty of the reanalysis is strongly dependent
Tuesday
on the observation density. FYRE Climate allows assessing for the first time
high-resolution observation-compatible multidecadal variations in both tem-
perature and precipitation over France.
Features and performances of FYRE Climate in terms of temperature and pre-
cipitation regimes, variability, and extremes are also compared over recent
overlapping periods with those of the Safran French reanalysis (Vidal et al.,
2010) and the Mescan European reanalysis (Soci et al., 2016).
Oral Presentations
TU.06 Toward a 35-years North American Pre-
cipitation and Surface Reanalysis: Evalua-
Wednesday
tion of a 5 years sample
Authors: Nicolas Gasset1,Vincent Fortin2, Marco Carrera2, Etienne Ga-
borit1, Milena Dimitrijevic2, Guy Roy1, Normand Gagnon1
1
Affiliations: Meteorological Service of Canada, Environment and Climate
Change Canada, Dorval, QC, Canada, 2Meteorological Research
Division, Environment and Climate Change Canada, Dorval,
QC, Canada
I n support of the International Watersheds Initiative (IWI) of the International
Oral Presentations
Joint Commission (IJC), a 35-years precipitation and surface reanalysis cover-
ing North America seamlessly at a 3-hours and 15-km resolution is currently
Thursday
being developed at the Canadian Meteorological Centre (CMC).
A deterministic reforecast / dynamical downscaling approach is followed
where a global reanalysis (ERA-Interim) is used as initial condition of a Nu-
merical Weather Prediction (NWP) system (the Global Environmental Multi-
scale model (GEM)) along with a coupling with precipitation and surface data
assimilation (i.e. the Canadian Precipitation Analysis (CaPA) and the Canadian
Land Data Assimilation System (CaLDAS)). While optimized to be more com-
putationally efficient in the context of a reforecast experiment, all systems
used are closely related to model versions and configurations currently run
Tuesday / Wednesday
Poster Presentations
operationally at CMC, meaning they have undergone a strict and thorough
validation procedure.
As a proof of concept and for evaluation purpose before achieving the
35-years reanalysis, the approach is evaluated for the years 2010-2014
against surface observations as well as currently available gridded reference
products such as NOAA’s Multisensor Precipitation Estimate (Stage IV) data-
set, the Global Precipitation Climatology Project (GPCP) precipitation data-
set, SMOS sattelite Surface moisture measurements, ERA-Interim and the
9International Symposium on Regional Reanalysis
newly released ERA5 reanalysis. A special attention is dedicated to the
evaluation of analysed variables, i.e. precipitation, surface/ground temperature
Oral Presentations
and moisture over the whole domain of interest. Finally, a preliminary evalu-
ation at the watershed level through hydrometeorological application is also
Tuesday
illustrated.
Results from this preliminary 5 years sample are very encouraging: the cou-
pled approach, i.e. GEM+CaPA/CaLDAS, always shows clear improvements
over classical reforecast / dynamical downscaling where surface observations
are present. Furthermore, results are inline or better than currently available
gridded reference products as well as the CMC operational approach that was
operated from 2012 to 2016.
Such a reanalysis is of prime interest for the whole North American hydro-
meteorology and land surface community, notably for transboundaries studies.
Oral Presentations
Among other things, it will allow for bias correction of current estimates and
Wednesday
forecasts, and help decision maker understand and communicate by how much
the current forecasted state of the system differs from the recent past.
TU.07 The 55 years UERRA Surface Re-Analysis
over Europe at 5.5km.
Authors: Eric Bazile1, Antoine Verrelle1, Patrick Le Moigne1, Francois
Besson1, Rachid Abida1, Camille Szczypta2, Cornel Soci3
1
Affiliations: Meteo-France, Toulouse, France, 2Mercator Océan, Toulouse,
France, 3ECMWF, Reading, UK
Oral Presentations
W ithin the European UERRA project ended end of 2017, a surface
Thursday
re-analysis was produced at high horizontal resolution (5.5km) over
Europe for the period 1961-2015. Furthermore, uncertainties were estimated
with an ensemble of 8 members at the same resolution. This database is avail-
able at ECMWF on the MARS archive.
This surface re-analysis done with the MESCAN-SURFEX system should
provide added value compared to global product such as ERA-Interim-Land
thanks to a finer horizontal grid (5.5km versus 80km), additional observations
and the use of a precipitation analysis (Soci et al. 2016). This precipitation
analysis done one a day improve significantly the precipitation field which is
Tuesday / Wednesday
Poster Presentations
one of the most essential variable for several applications: hydrological model
for water management or snow pack evolution over mountains.
After a brief description of the MESCAN-SURFEX system and the output data
available, we will present the results and comparison for the 55 years period
with independent observations and global re-analysis. A focus will be done
on the evaluation of precipitation and the snow depth in particular over the
mountainous area. The question of uncertainties and the impact of the obser-
vation network will be addressed.
10ISRR
2018
Bonn
Germany
Additional information about the UERRA project can be found at http://www.
uerra.eu
Oral Presentations
The research leading to these results has received funding from the Europe-
an Union, Seventh Framework Programme (FP7-SPACE-2013-1) under grant
Tuesday
agreement no 607193.
TU.08 Reanalysis for wind and solar electricity
simulations: challenges and lessons learned
in the Renewables.ninja project (Invited
talk)
Authors: Stefan Pfenninger1, Iain Staffell2
Oral Presentations
1
Affiliations: ETH Zürich, Switzerland, 2Imperial College London, UK
Wednesday
T he Renewables.ninja project uses meteorological reanalysis data to de-
velop freely available worldwide simulations of solar and wind electricity
generation. This talk will discuss the approaches used for simulation and bias
correction and show some recent improvements made based on user feed-
back, including directly providing energy-relevant weather variables and more
pre-aggregated data on different spatial scales.We will then show a preliminary
comparison using MERRA-2 (the primary source of data currently used), ERA-
5, and COSMO-REA6, discussing the advantages and disadvantages of regional
reanalysis data for our applications. The talk will conclude by highlighting some
recent applications of our data and give an outlook of future developments.
Oral Presentations
TU.09 Impact of different weather data sets on
Thursday
the results of a power system model
Authors: Gerda Deac1, Katja Franke1
1
Affiliation: Fraunhofer Institute for Systems and Innovation Research ISI
W ith rising shares of renewable energies in the power system, weather
data become a crucial input in the energy system modelling. We use
weather data in our energy system optimization model Enertile, which in-
Tuesday / Wednesday
Poster Presentations
cludes a potential analysis for wind and solar power for Europe, North Africa
and the Middle East. Different weather data sources with different spatial
resolution affect the calculated potential of renewable energies and thus the
result of the energy system optimization as e.g. power plant expansions. We
analyze the impact of different weather datasets on modelling results focusing
on differences in renewable potentials and hourly performance of renew-
ables in a power system model. Based on the analysis, we provide an insight on
the importance of weather data in energy system modelling.
11International Symposium on Regional Reanalysis
TU.10 Assessment of complementarity between
wind power and photovoltaic installations
Oral Presentations
to supply residential electric demand in
Germany and Czech Republic
Tuesday
Authors: Luis Ramirez Camargo1, Katharina Gruber1,2, Felix Nitsch1,2
1
Affiliations: Institute for Applied Informatics, Technische Hochschule Deg-
gendorf, Freyung, Germany, 2Institute for Sustainable Economic
Development, University of Natural Resources and Life Scienc-
es,Vienna, Austria
T he electricity self-sufficiency of individual single-family houses (SFH) in
Oral Presentations
Central Europe can be achieved with combinations of photovoltaic (PV)
and battery systems. System sizing strongly depends on particular weather
Wednesday
conditions but even in privileged locations technical requirements are high;
the total system costs would be far beyond average household’s economic
possibilities and without a grid connection more than 60% of the total PV
generation would have to be curtailed. In order to reduce curtailment and
investment costs of PV and battery systems, the present work evaluates the
potential of combinations of these with wind power for supply of electricity
for groups of SFHs. This study relies on 20 years of hourly solar radiation,
wind speed and temperature data from the COSMO-REA6 regional reanalysis,
satellite imagery derived snow cover data, land use and population data as
well as a techno-economical optimization model to size renewable hybrid
Oral Presentations
energy generation systems. Maps showing the required number of small wind
turbines (10.5 kW), PV and battery systems sizes for clusters of self-suffi-
Thursday
cient SFHs across Germany and Czech Republic are generated for multiple
technical and weather dependent scenarios. Results are discussed considering
studies of complementarity between wind and solar power as well as previous
work on self-sufficient SFHs in Central Europe.
Tuesday / Wednesday
Poster Presentations
12Oral Presentations Oral Presentations Oral Presentations Poster Presentations
Tuesday Wednesday Thursday Tuesday / Wednesday
13
ISRR
Germany
Bonn
2018Oral Presentations Wednesday - 18 July 2018
ISRR
2018
Bonn
Germany
WE.01 Estimating the value of regional reanalyses
from the UERRA inter-comparison
Oral Presentations
Authors: Andrea K. Kaiser-Weiss1, Deborah Niermann1, Michael
Tuesday
Borsche1, Cristian Lussana2, Francesco Isotta3, Else van Besse-
laar4, Gerard van der Schrier4, Frank Kaspar1, Per Unden5
1
Affiliations: Deutscher Wetterdienst (DWD), Germany, 2Meteorologisk
Institutt (MI), Norway, 3Eidgenoessisches Departement Des
Innern (EDI), Meteoswiss, Switzerland, 4Koninklijk Nederlands
Meteorologisch Instituut (KNMI), Netherlands, 5Sveriges Mete-
orologiska Och Hydrologiska Institut (SMHI), Sweden
R egional reanalyses provide commonly requested meteorological parame-
Oral Presentations
ters like wind speed, radiation, temperature and precipitation. Here un-
Wednesday
certainty estimates are discussed resulting from the inter-comparison of three
European regional reanalyses (based on numerical weather models COSMO,
UM and HARMONIE-ALADIN), several downscaling products (based on MES-
CAN) and the observation-based E-OBS dataset recently produced within the
EU FP7 project Uncertainties in Ensembles of Regional ReAnalysis (UERRA).
The performance of the regional reanalyses depends on the location, the sea-
son and on the spatio-temporal scale of interest. Most pronounced is the value
of reanalyses in locations of data sparse areas, where the numerical weather
models are superior in transporting information compared to the traditional
gridding procedures. For wind speed at heights relevant for wind energy, an
added value over ERA-Interim could be demonstrated for all UERRA reanal-
Oral Presentations
yses products. Radiation fields show a model-dependent bias against satellite
data. Station temperatures were generally found in agreement, however with
Thursday
a bias becoming severe for the (moderately) extreme values, with potential
pitfalls for threshold applications such as climate indices. Comparisons of the
precipitation fields in different areas of Europe demonstrate that the various
reanalysis excel in different regions, coinciding with the respective areas the
models are applied operationally. The multi-model ensemble was found to
provide better uncertainty estimates than the ensemble of one model alone.
WE.02 Vertical wind profile: Assessment of re-
Tuesday / Wednesday
Poster Presentations
gional reanalyses.
Authors: Christopher W. Frank1,2, Sabrina Wahl1,3, Jan D. Keller1,4, Bern-
hard Pospichal2, Andreas Hense3, Susanne Crewell2
1
Affiliations: Hans-Ertel-Centre for Weather Research, Climate Monitoring
and Diagnostics, Germany, 2Institute of Geophysics and Me-
teorology, University of Cologne, Germany, 3Meteorological
Institute, University Bonn, Germany, 4Deutscher Wetterdienst,
15International Symposium on Regional Reanalysis
Offenbach, Germany
P
Oral Presentations
ublic safety, renewable energy, aeronautical meteorology: For a variety of
applications the vertical wind profile needs to be known as accurately
Tuesday
as possible at any time and for any location. Knowing the true vertical
wind profile would improve numerical weather prediction, would enhance
renewable energy production estimates, and reduce aviation risks. Unfortu-
nately, wind profile measurements provided by weather masts or wind pro-
filer systems are spatially rare and limited to the lower atmosphere. Another
source of vertical wind profiles are reanalyses. Reanalyses combine numeri-
cal weather prediction models and observations to provide a best estimate
of the atmospheric flow in the past. Reanalyses have the advantage to provide
the wind speed on a homogeneous grid in space and time. Recently new high
Oral Presentations
resolution regional reanalyses have been developed to better address the
local scale, however, their quality still needs to be quantified.
Wednesday
Here we investigate the capabilities of regional reanalyses developed within
the Hans-Ertel Centre for Weather Research to reproduce realistic wind
profiles. The high resolution reanalyses COSMO-REA6 (6 km horizontal res-
olution, 40 vertical layer) and COSMO-REA2 (2 km horizontal resolution, 50
vertical layer) are both based on the NWP model COSMO. REA6 covers the
EURO-CORDEX region and REA2 an extended COSMO-DE domain over
central Europe. The coarser reanalysis is available from 1995 to 2015 and
the finer from 2007 to 2013.
The central part of this contribution is to quantify the uncertainty of wind
profiles provided by regional reanalyses. Therefore, reanalyses profiles of the
Oral Presentations
horizontal wind speed are compared to weather mast measurements. A first
assessment considers annual and daily cycle comparisons. Since wind profiles
Thursday
depend primarily on atmospheric stability ,the performance is quantified for
selected stability conditions. For a realistic assessment of the results, the local
conditions around the weather masts are taken into account.
WE.03 What do we gain from the regional reanal-
ysis products - a perspektive from multi-
variate analysis of PBL Profiles
Tuesday / Wednesday
Poster Presentations
Authors: Ronny Petrik1, Burkhardt Rockel1, Beate Geyer1
1
Affiliation: Helmholtz-Zentrum Geesthacht
E nergy system models need to have reliable and robust weather input data
to simulate over many years the production of energy and the flow with-
in the power grid. For weather data we rely on reanalysis/hindcast products
which deliver various data of interest in 3D-space and time.
With respect to renewables the main challenge is that all information need-
16ISRR
2018
Bonn
Germany
ed is a) related to variables resulting from strong nonlinear interactions in
the atmosphere (as for radiation) or is b) related to levels located above
Oral Presentations
the surface (as for wind). In both cases an assimilation within a reanalysis
framework is hard to realize. Therefore, not only reanalysis products but also
Tuesday
hindcasts based on large-scale constraints and pure hindcasts are analyzed for
their quality with a focus on the planetary boundary layer.
Different measurement towers and remote sensing devices spread over
Europe are taken as reference data for model evaluation. The profiles
of temperature, humidity and wind speed are investigated and the differ-
ent reanalysis products are intercompared. The intercomparison reveals the
strength and weaknesses of each member. In addition, it is highlighted where
the reanalysis systematically reduces model deficits compared to a hindcast,
but also which model errors survive in the reanalysis. The verification aims to
Oral Presentations
distinguish between different atmospheric regimes and furthermore assesses
Wednesday
the temporal consistency of lower tropospheric variability by means of cor-
relation and ramp statistics.
Finally, the turbulence intensity is analyzed because it originates from a com-
plex interaction between the different variables mentioned before. Although
the turbulence statistics is not explicitly resolved by the recent reanalysis
products, the effect of turbulent fluxes on the mean atmospheric fields can
be parameterized. It is shown to which extent the models are able to rep-
resent the mean characteristic of deviations from mean wind field.
WE.04 Intercomparison of the representation of
Oral Presentations
climate in global and regional reanalysis
Thursday
Authors: Jan D. Keller1,2, Sabrina Wahl1,3
1
Affiliations: Hans-Ertel-Centre for Weather Research, Climate Monitoring
and Diagnostics, Germany, 2Deutscher Wetterdienst, Offenbach,
Germany, 3University of Bonn, Germany
T he importance of reanalyses as a basis for numerous applications in geo-
sciences, economics and other fields has significantly grown over the last
decade. Therefore, a comprehensive evaluation and intercomparison of reanal-
ysis data sets is essential. Here, we quantify the added value of regional atmo-
Tuesday / Wednesday
Poster Presentations
spheric reanalyses (such as COSMO-REA6 or the HARMONIE reanalysis)
compared to the global (forcing) reanalysis products (e.g., ERA-Interim,
MERRA-2) with respect to monitoring climate and its variability. To this end,
we investigate the representation of climate indices in the reanalyses in
comparison to indices derived from observations. The set of climate indices
in consideration comprises simple measures such as number of summer days
or frost days but also aggregated indices, e.g., heating / cooling degree days. In
order to assess the quality of the climate representation and thereby the
17International Symposium on Regional Reanalysis
added value of regional data sets, we also quantitatively verify the indices
derived from reanalyses with independent observations.
Oral Presentations
WE.06 European high resolution Regional Re-
Tuesday
analyses in UERRA and the Copernicus
Climate Change Service (C3S)
Authors: Per Unden1, Semjon Schimanke1, Patrick Le Moigne2, Eric
Bazile2, Martin Ridal1, Esbjorn Olsson1, Lars Berggren1, Roger
Randriamampianina3
1
Affiliations: Swedish Meteorological and Hydrological Institute (SMHI),
Norrkoping, Sweden, 2Meteo-France, Toulouse, France, 3Met
Oral Presentations
Norway, Oslo, Norway
Wednesday
W ithin the EU FP7 SPACE programme regional European Reanalyses sys-
tems were developed and produced for multi-decadal periods by 4
partners (SMHI, Met Office, Meteo-France and University of Bonn/DWD).
SMHI coordinated the project and produced an 11 km grid resolution Eu-
ropean Reanalysis from 1961. It has continued to near real time within
the Copernicus programme (C3S) where SMHI is contracted to develop
and produce a regional Reanalysis for Europe within the Copernicus climate
change services (PRECISE). Meteo-France and Met Norway are also working
in the Project.
The UERRA system run by SMHI is with the ALADIN model in the HARMO-
Oral Presentations
NIE environment developed within the HIRLAM modelling consortium. The
horizontal grid is at 11 km and there are 65 model levels up to 10 hPa. Con-
Thursday
ventional observations from ECMWF MARS are used with complemented
with extra surface observations from some countries. The global ERA-Inter-
im (and ERA-40) reanalyses provide lateral boundaries and large scale forc-
ing through an extra constraint in the 3D-VAR analysis. The reanalysis data
are freely available through the MARS data services at ECMWF (in GRIB-2).
The 11 km resolution provides much more detail and realistic local features
than is possible in the about 78 km global reanalysis. The UERRA partners
have carried out an extensive evaluation of the quality and uncertainties of
the SMHI reanalysis as well as of the Met Office and University of Bonn ones
Tuesday / Wednesday
Poster Presentations
and the Meteo-France surface parameter reanalysis. There both benefits and
uncertainties which vary over region and with horizontal scales.
The PRECISE system for the C3S European Reanalysis is being developed and
contains several enhancements. The resolution is roughly doubled, the cycling
in time will be 3 hours instead of 6 and most importantly, there will be and
ensemble assimilation coupled with the deterministic high resolution run.
18ISRR
2018
Bonn
Germany
WE.07 The Arctic Regional Reanalysis of the Co-
pernicus Climate Change Service
Oral Presentations
Authors: Harald Schyberg1, Heiner Kornich2, Kristian Pagh Nielsen3,
Tuesday
Roger Randriamampianina1, Xiaohua Yang3
1
Affiliations: Norwegian Meteorological Institute, 2Swedish Meteorological
and Hydrological Institute, 3Danish Meteorological Institute
W e will present status and plans for the Arctic regional reanalysis of the
Copernicus Climate Change Service (C3S), C3S project 322 Lot 2. The
project aims to produce an Arctic regional reanalysis over two subdomains of
interest for change processes and economic activities. The reanalysis will cover
the period 1997 - 2021 with a horizontal resolution of 2,5 km. Additionally a
Oral Presentations
proof-of-concept for a pan-Arctic reanalysis will be provided for a period of
Wednesday
1-year.
The system to be used is based on the HARMONIE-AROME Numerical Weather
Prediction (NWP) system, with additions and configuration choices for reanal-
ysis purposes with that system. Global reanalysis data from ERA5 will be used
for lateral boundaries.
Development activities connected to the system includes work on assimila-
tion setup, 3D-Var background error statistics and uncertainty estimation.
The Arctic reanalysis will add value versus the global reanalysis by providing
higher-resolution and by using regional data not used there. The upper air
assimilation will use conventional observations and, since there are gaps in
the conventional observing system, will put emphasis of using satellite data-
Oral Presentations
sets which have good coverage in the Arctic. This will include important parts
of the satellite observing system such as microwave and infrared radiances,
Thursday
atmospheric motion vectors, scatterometer winds and radio occultation data.
Handling of “cold surfaces” in the surface scheme, such as snow, sea ice and
glaciers, which are important in the Arctic, will also receive special attention
with the aim to give a better representation than in the global reanalysis.
A main milestone will be the start of the reanalysis production, which is
planned to take place in April 2019, and the reanalysis dataset will then be
continuously updated to provide a complete 24 years time series before the
end of the project.
Tuesday / Wednesday
Poster Presentations
WE.08 SMHIs applications of UERRAs and Coper-
nicus regional reanalysis for Europe
Authors: Semjon Schimanke1, Ludvig Isaksson1, Lisette Edvinson1, Per
Unden1, Martin Ridal1
1
Affiliations: Swedish Meteorological and Hydrological Institute (SMHI),
Norrkoping
19International Symposium on Regional Reanalysis
W e are going to present how the regional reanalysis (RRA) from UERRA
and its follow-up service by Copernicus is used at SMHI.
Oral Presentations
Based on the pre-operational FP7 project UERRA RRA data is available for
Europe for the period 1961-2015. The same model system is now used in
Tuesday
the operational C3S service Copernicus regional reanalysis for Europe, where
the data set is updated monthly. Herewith, users are offered a continuous and
homogeneous data set for the period 1961 to close to real time. In our presen-
tation we will show how the RRA is used at SMHI in the moment and which
future plans we have for the usage of the RRA. Examples of usage are:
1. Correction of historical observations: Based on comparisons between the
RRA and SMHIs observational data base (MORA) our climatologist man-
age to find errors in MORA, which they did not discover earlier. Here-
with, the RRA helps to improve the quality of SMHIs data base, which will
Oral Presentations
even have a positive impact on future reanalysis products
Wednesday
2. Season map: SMHI produces daily an overview showing the prevailing
season of the year following SMHIs definition. Here, the RRA data is
used to show climatological arrival dates of the seasons as well as for
constructing a historical data base with arrival dates over the years
1961-2017
3. Forcing data for regional ocean models: The RRA parameters 2m-tem-
perature, humidity, wind, SLP, precipitation and radiation are used to
drive hindcast simulations with the regional ocean model NEMO-Nordic.
Even other examples might be discussed in the presentation.
Moreover, we will critically review the quality of the RRA data set. At SMHI,
Oral Presentations
we had even more ideas and plans on how to use the RRA. However, during
the preparation of different applications, we discovered certain shortcomings
Thursday
in the RRA, which prevented us from setting up corresponding services. Exam-
ples will be given in the presentation.
WE.09 Land Surface Hydrology in the European
High-resolution Regional Reanalysis UER-
RA
Authors: Patrick Le Moigne1, Antoine Verrelle1, Camille Szczypta2, Eric
Bazile1
Tuesday / Wednesday
Poster Presentations
1
Affiliations: Meteo-France, Toulouse, France, 2Mercator Ocean, Toulouse,
France
T he UERRA project (2014-2017), financed by the European Union under
its 7th Framework Programme SPACE, aimed at building a high resolution
(5.5km) reanalysis of surface essential climate variables (ECV) over Europe,
covering the period 1961-2015, and at estimating associated uncertainties.
The atmospheric UERRA reanalysis relies on the HARMONIE system and
20ISRR
2018
Bonn
Germany
the ALADIN model at 11km horizontal resolution. A downscaling at 5.5km is
performed to derive wind and the downwards radiative fluxes, and to provide
Oral Presentations
background fields for the MESCAN surface analysis of 2m temperature and
relative humidity and 24h accumulated precipitation. These mesoscale fields
Tuesday
are then used to drive hydro-meteorological off-line simulations based on
the coupled SURFEX-CTRIP system, where SURFEX is the land surface
model developed and used at Meteo-France for NWP and climate applications,
and CTRIP is the river routing model used for climate applications at Me-
teo-France. Such a system provides not only ECVs, such as soil moisture, snow
depth, etc., but also computes the various components of the water cycle:
precipitation, surface runoff, soil infiltration and water storage, as well as the
time evolution of rivers discharge and aquifers height.
A description of the MESCAN-SURFEX-CTRIP system will be done first. A
Oral Presentations
detailed analysis of the water cycle components climatology over the 1961-
Wednesday
2015 period over Europe will be done and comparisons with independent
observations of river discharges will be discussed.
Additional information about the UERRA project can be found at http://www.
uerra.eu
The research leading to these results has received funding from the Europe-
an Union, Seventh Framework Programme (FP7-SPACE-2013-1) under grant
agreement no 607193.
WE.10 Continental-scale high resolution terres-
trial hydrologic modeling over Europe us-
Oral Presentations
ing the COSMO regional reanalysis
Thursday
Authors: Bibi S. Naz1,2, Stefan Kollet1,2, Carsten Montzka1, Anne Springer3,
Klaus Goergen1,2, Carina Furusho1,2
Affiliations: 1
Institute of Bio- and Geosciences (Agrosphere, IBG-3), Re-
search Centre Juelich, Germany, 2Centre for High-Performance
Scientific Computing in Terrestrial Systems, Geoverbund ABC/J,
Juelich, Germany, 3Institute of Geodesy and Geoinformation,
Bonn University, Germany
M eteorological forcing is critical for hydrological modeling as it has a strong
Tuesday / Wednesday
Poster Presentations
impact on the accuracy of simulated hydrological states and fluxes. How-
ever, meteorological observations over large spatial scales at high resolution
are often lacking.The use of high-resolution regional reanalysis data which pro-
vide regionally consistent dataset of meteorological variables may be a valuable
approach to drive terrestrial hydrologic models for studying water resources
planning, impact assessments, and flood forecasting. This study investigates the
use of high resolution reanalysis datasets for hydrological modeling and its ef-
fects on hydrological extremes. The community land model (CLM 3.5) was set
21International Symposium on Regional Reanalysis
up over Europe at 3km resolution and forced with the COSMO-REA6 from
Hans-Ertel Centre for Weather Research (HErZ) for a time period of 2000
Oral Presentations
to 2014. The performance of the model was analyzed through comparisons
with independent observations including satellite soil moisture, gridded ob-
Tuesday
servational-based runoff, GRDC river discharge and total water storage from
GRACE satellite. Overall, our results indicate that in regions with high spa-
tial variability of precipitation, high-resolution reanalysis performs better than
coarse-resolution reanalysis dataset and suggest that reanalysis products are
useful for hydrological modeling over large scales.
WE.11 Reproduction of storms over the North
Sea and the Baltic with the regional
Oral Presentations
reanalysis COSMO-REA6
Wednesday
Authors: Natacha Fery1, Birger Tinz1, Lydia Gates1
1
Affiliation: Deutscher Wetterdienst, Marine Climate Monitoring, Hamburg,
Germany
I n this study, we assess how severe storms in the North Sea and the Baltic
between 1994 and 2015 are reproduced in the high resolution reanalysis
COSMO-REA6 (Deutscher Wetterdienst, University of Bonn). The analysis
was carried out for the storms Anatol (12/1999), Britta (11/2006), Kyrill
(01/2007), Tilo (11/2007), Christian (10/2013) and Xaver (12/2013), storms
strongly varying in intensity and duration. The results of the regional reanal-
Oral Presentations
ysis were compared to the quality controlled and mast corrected measure-
ments of the research platforms FINO1, 2 and 3 in the North and the Baltic
Thursday
Sea. The comparison for stormy conditions included wind speed, wind direc-
tion and pressure at the surface and at several heights along the masts of the
platforms (from 20 m to 100 m depending on the location). To provide infor-
mation about the added value of using regional reanalysis as compared to the
global reanalysis in reproducing severe storms over sea, additional compari-
sons of the 10 m wind and the sea level pressure between COSMO-REA6
and the global atmospheric reanalyses ERA-40, ERAInterim and 20CRv2c
are provided. Preliminary results suggest that regional reanalyses are better
suited to reproduce short, fast moving and strong storms. Both the regional
Tuesday / Wednesday
Poster Presentations
and the global reanalyses are able to reproduce the atmospheric conditions
of storms characterized by a large duration and mostly associated to storm
surges. This work is part of the project EXTREMENESS, funded by the Ger-
man Federal Ministry of Education and Research that aims at investigating
extreme storm surges and their consequences along the German North Sea
coast. It finds application in the field of engineering design of coastal protec-
tion infrastructure that relies on accurate reproduction of storms, especially
the storm surges that may lead to significant damages along the coasts.
22ISRR
2018
Bonn
Germany
WE.12 Using the Australian regional atmospheric
reanalysis to assess climate-related haz-
Oral Presentations
ards
Tuesday
Authors: Chun-Hsu Su1, Nathan Eizenberg1, Christopher J. White2,3,
Stuart Matthews4, Paul Fox-Hughes1,2, Doerte Jakob1, Samuel
Sauvage1, Mitchell Black1, Robert Fawcett1, Robert Smalley1,
Peter Steinle1
Affiliations: 1
Bureau of Meteorology, Australia, 2Antarctic Climate and
Ecosystems Cooperative Research Centre, Hobart, Australia,
3
Department of Civil and Environmental Engineering, Univer-
sity of Strathclyde, Glasgow, UK, 4New South Wales Rural Fire
Oral Presentations
Service, Lidcombe, NSW, Australia
T
Wednesday
he Bureau of Meteorology Atmospheric high-resolution Regional Reanaly-
sis for Australia (BARRA, http://www.bom.gov.au/research/projects/reanal-
ysis) is the first atmospheric reanalysis produced by the Bureau for the Austra-
lian region. It is co-funded by several state fire agencies and state government
agencies. When completed in 2019, BARRA will provide over 25 years of
hourly reconstructions of past atmospheric conditions from 1990 onwards.
BARRA’s 12-km regional reanalysis (BARRA-R) over the Australia and New
Zealand regions is bounded by the ERA-Interim reanalysis. Importantly, BAR-
RA includes multiple convection-permitting (1.5-km) dynamical downscaling
analyses (BARRA-xx) over smaller sub-domains, nested within BARRA-R, to
provide additional fine-scale detail.
Oral Presentations
By providing additional fine-scale detail from the surface to the lower strato-
sphere, beyond what can be obtained by coarser-scale global reanalyses or
Thursday
global climate models, BARRA supports high-quality assessments of weather
and climate related risks and is well suited for the study of high-impact histor-
ical events. BARRA will provide information on the multi-decadal time scales
required for assessing hazards and risks in the present climate. We anticipate
that in time BARRA and its successors will be an important supplement to
existing climate information.
BARRA is used by fire agencies for developing a new climatology of fire
diagnostics and to drive fire predictive model simulations for representative
Tuesday / Wednesday
Poster Presentations
fire weather events. BARRA has also become a critical data set for refining
the threshold values of the new rating categories for an Australian National
Fire Danger Ratings System (NFDRS), and permitting an understanding of the
behaviour of the NFDRS over an extended period. Significant variability in the
density of observational networks across Australia and the lack of high-qual-
ity observation networks for solar, wind and sub-surface parameters provide
impetus for using BARRA to (i) create climatologies for variables such as
wind, radiation, and soil moisture, (ii) enhance the Bureau’s existing spatially
23International Symposium on Regional Reanalysis
interpolated gridded climate data sets for precipitation and temperature, and
(iii) provide tailored services to satisfy requests from both public and private
Oral Presentations
sectors. In this presentation, we discuss these growing uses of BARRA in the
areas of data provision and services, disaster and risk management.
Tuesday
WE.13 Remote sensing of global snow cover (In-
vited talk)
Authors: Andreas J. Dietz1, Claudia Künzer1
1
Affiliation: German Aerospace Center (DLR), German Remote Sensing
Data Center (DFD), Wessling, Germany
S
Oral Presentations
now cover is an important variable for water availability, the radiation bud-
get, glaciers, flora and fauna, and may cause natural disasters such as ava-
Wednesday
lanches or floods. In many countries, snow is an important source of freshwa-
ter for reservoirs and the subsequent production of electricity. Climate change
is affecting the global snow cover distribution, extent, and mass, influencing
all the aforementioned parameters. It is therefore important to monitor the
developments and changes to be able to detect possible trends and future
impacts of changing snow cover on our environment.
Remote Sensing offers an ideal data source to detect global snow cover with
both high temporal and spatial resolution. Observations from medium resolu-
tion sensors such as AVHRR, MODIS, or Sentinel-3 provide time series of daily
data, which can be processed and classified to calculate the snow cover extent
Oral Presentations
on a global scale. Polar night, cloud cover, and complex terrain in mountain
regions can cause data gaps and classification uncertainties. To overcome these
Thursday
problems, several techniques can be applied to the data in order to provide
a consistent time series of daily snow cover information, which can further
be processed to derive parameters such as snow cover duration per year or
beginning/end of snow cover season. These parameters can be analysed to
quantify the effects of climate change on snow cover on local, regional, and
global scales.
The presentation will outline the different steps in order to obtain snow cover
information from remote sensing data and how these snow cover parameters
can further be processed. Several examples will be presented for possible ap-
Tuesday / Wednesday
Poster Presentations
plications, and the opportunity of combining the results from remote sensing
with those obtained by modelling/re-analysis will also be discussed.
WE.14 Data assimilation of urban weather obser-
vations in WRF to model the urban cli-
mate of Amsterdam.
24ISRR
2018
Bonn
Germany
Authors: Sytse Koopmans1, Ronald van Haren2, Gert-Jan Steeneveld1,
Natalie Theeuwes3, Remko Uijlenhoet1, Albert A.M Holtslag1
Oral Presentations
1
Affiliations: Wageningen University and Research, the Netherlands,
2
Netherlands eScience Center, the Netherlands, 3University of
Tuesday
Reading, United Kingdom
O ngoing world-wide climate change and urbanization illustrate the need
to understand urban hydrometeorology and its implications for human
thermal comfort and water management. Numerical weather prediction
models can assist to understand these issues, as they progress increasingly
towards finer scales. With high model resolutions (grid spacing of 100m),
effective representation of cities becomes crucial. The complex structures of
cities, configuration of buildings, streets and scattered vegetation, require a
Oral Presentations
different modelling approach than the homogeneous rural surroundings. The
Wednesday
current urban canopy-layer schemes account for these city specific character-
istics, but differ substantially amongst each other due to uncertainty in land use
parameters and incomplete physical understanding. Therefore, the hindcasting
of the urban environment needs improvement.
In this study, we improve the WRF (Weather Research and Forecasting) meso-
scale model performance by incorporating observations of a variety of sourc-
es using data assimilation (WRF-3DVAR) and nudging techniques on a resolu-
tion up to 100 meter. Data assimilation aims to accurately describe the most
probable atmospheric state by steering the model fields in the direction
of the observations. Assimilated observations consists of WMO synoptic
weather observations, volume radar data and urban weather observations
Oral Presentations
recorded by hobby meteorologists.
Thursday
Specific to urban boundary layers, a novel approach has been developed
to nudge modelled urban canyon temperatures with quality controlled ur-
ban weather observations. Adjusting the urban fabric accordingly is crucial,
because of the large heat storage within urban canopies. The road and wall
layers of the urban canopy are adjusted depending on the bulk heat transfer
coefficient and urban geometry. In addition, the representation of the anthro-
pogenic heat release within the urban canopy model SLUCM is improved
by incorporating this flux predominantly into the canyon instead of the first
model layer above the canyon.
Tuesday / Wednesday
Poster Presentations
The subsequent data assimilation steps are evaluated for hindcasts of July
2014 for the Netherlands. July 2014 is characterized by both a warm dry
period and two days with extreme precipitation (more than 100mm
in two days). The largest improvement is made by assimilating the air
temperature, dew point temperature and pressure from WMO synoptic
stations. Assimilating additional radar data, slightly improves the location of
the precipitation indicated by the fraction skill score. The data assimilation of
urban weather stations, lastly, reduces the cold biases within the urban canopy
25International Symposium on Regional Reanalysis
which appeared in WRF.
Our final goal is to create a 15-year climatological urban re-analysis data
Oral Presentations
archive of (hydro)meteorological variables which is named ERA-urban. This
will enable us to trace trends in thermal comfort and extreme precipitation.
Tuesday
WE.15 Towards a Convective Scale Atmospheric
Reanalysis including a Soil-Vegetation-At-
mosphere-Transfer-Model
Authors: Clarissa Figura1, Insa Thiele-Eich1, Jan D. Keller2,3, Wolfgang
Kurtz4, Clemens Simmer1, Andreas Hense1
1
Affiliations: Meteorological Institute, University of Bonn, Germany,
Oral Presentations
2
Hans-Ertel-Centre for Weather Research, Bonn, Germany,
Wednesday
3
Deutscher Wetterdienst, Offenbach, Germany, 4Institute of Bio-
and Geosciences (Agrosphere, IBG-3), Research Centre Jülich,
Germany
R eanalyses data sets are not only of value in the atmospheric sciences, but
also in other disciplines such as hydrology or agriculture, where they can
be used in numerous applications. Current global reanalyses, albeit being a
valuable source of information, lack a comprehensive representation of the
terrestrial system as a whole, including land surface and subsurface espe-
cially if data for the meso β/γ scale are required. For instance, diurnal to
seasonal water cycles in subcontinental river catchments are often not well
Oral Presentations
represented.
Recent studies showed that employing global reanalyses as forcing for land
Thursday
surface model simulations of specific river basins, with higher grid spacing, lead
to an improvement in the representation of soil moisture and precipitation.
The next step is to allow a feedback between the improved land surface and
the atmosphere to take place within the reanalysis system.
This study focuses on a retrospective analysis of a short summer period for
a river catchment in Germany. The setup includes a limited area model at
horizontal resolutions of 1 km for the atmosphere and 500 m for the
soil. The simulations are based on the Terrestrial Systems Modeling Platform
(TerrSysMP) which consists of three model components, the atmospheric
Tuesday / Wednesday
Poster Presentations
model COSMO, the surface model CLM3.5 (Community Land Model) and
the 3-dimensional hydrological model ParFlow, connected via an exter-
nal coupler (OASIS3), for the exchange of relevant state vectors. During
the simulation, rain radar reflectivities are continuously assimilated with the
KENDA framework, which is the local ensemble transform Kalman filter
(LETKF) implementation, operationally used at the German Meteorological
Service (DWD). The results are analysed with a focus on precipitation and
soil moisture in order to evaluate the added value of using a fully coupled
26You can also read
