Geoscience Reference
In-Depth Information
Concerning soil moisture, most NWP centres use a 1-D Optimal Interpolation analysis
to initialise soil moisture, based on a dedicated screen-level parameters analysis. Both the
German meteorological service and ECMWF use an Extended Kalman Filter soil moisture
analysis in operations. The Extended Kalman Filter soil moisture analysis is based on a
dedicated screen-level parameters analysis. Whereas the 1D Optimal Interpolation soil
moisture analysis uses screen-level analysis increments as input of the soil moisture
analysis, the Extended Kalman Filter, as implemented at ECMWF, uses analysed screen-
level fields as input observations of the soil moisture analysis. ECMWF experiments
showed that the Extended Kalman Filter soil moisture analysis consistently reduces, by
5 mm per month at global scale, the soil moisture increments compared to the Optimal
Interpolation, and it slightly improves both soil moisture and screen-level parameters
analyses and forecasts. The ECMWF soil moisture analysis evaluation against ground
measurements from the SMOSMANIA network showed an improved correlation from 0.78
for the Optimal Interpolation to 0.84 for the Extended Kalman Filter soil moisture analysis.
The Extended Kalman Filter analysis also makes it possible to combine screen-level
parameters and satellite data, such as ASCAT or SMOS, to analyse soil moisture. Previous
results with ASCAT data assimilation were discussed in this paper. ECMWF results
showed a neutral impact of ASCAT data assimilation in the Extended Kalman Filter on
both soil moisture and screen-level parameters. However, recent improvements in the
ASCAT soil moisture products and in bias correction are expected to improve the impact
of using ASCAT soil moisture data. In contrast to other centres, which mainly use Optimal
Interpolation or Extended Kalman Filter approaches, the UKMO, soil moisture analysis
relies on a simple nudging scheme. ASCAT soil moisture data assimilation was shown to
have a positive impact on the screen-level parameters forecast at the UKMO, leading to
operational ASCAT soil moisture assimilation from 2010. Developments are ongoing at
the UKMO to replace their current nudging scheme by an Extended Kalman Filter
approach which will open possibilities to combine different types of observations in their
soil moisture analysis.
Kalman Filter-based land surface analysis systems (Ensemble or Extended), used in
several NWP centres, either in research or operationally, open a wide range of further
development possibilities, including exploiting new satellite surface data and products for
the assimilation of soil moisture (e.g., SMOS or the future SMAP). At ECMWF an
extension of the Extended Kalman Filter to analyse additional variables, such as snow
temperature, snow mass and vegetation parameters, is planned for investigation in the near
future.
Acknowledgments The authors thank two anonymous reviewers for their careful review of the manuscript
and their helpful suggestions.
References
Albergel C, de Rosnay P, Gruhier C, Sabater JM, Hasenauer S, Isaksen L, Kerr Y, Wagner W (2012)
Evaluation of remotely sensed and modelled soil moisture products using global ground-based in-situ
observations. Remote Sens Environ 18:215-226. doi:
10.1016/j.rse.2011.11.017
Balsamo G, Mahfouf JF, B´lair S, Deblonde G (2007) A land data assimilation system for soil moisture and
temperature: an information content study. J Hydrometeorol 8:1225-1242. doi:
10.1175/2007JHM819.1
Balsamo G, Viterbo P, Beljaars A, van den Hurk B, Hirsch M, Betts A, Scipal K (2009) A revised hydrology
for the ECMWF model: verification from field site to terrestrial water storage and impact in the
Integrated Forecast System. J Hydrometeorol 10:623-643
