TREND ANALYSES OF A GLOBAL SOIL MOISTURE TIME SERIES DERIVED FROM
ERS-1/-2 SCATTEROMETER DATA:
FLOODS, DROUGHTS AND LONG TERM CHANGES
C. Kuenzer 3 ’ *, Z. Bartalis b , M. Schmidt 3 , D. Zhao 3 , W. Wagner b
a German Remote Sensing Data Center, DFD, of the German Aerospace Center, DLR, Postfach 1116, 82230 Wessling,
Germany - claudia.kuenzer@dlr.de (at time of preparation of abstract still at IPF of Vienna University of Technology)
b Institute of Photogrammetry and Remote Sensing, IPF, Vienna University of Technology, Gusshausstr. 27-29, A-
1040 Austria - zb@ipf.tuwien.ac.at, ww@ipf.tuwien.ac.at
c Key Laboratory of Regional Climate-Environment for Temperate East Asia, Institute of Atmospheric Physics, Chinese Academy
of Sciences, CAS, Beijing 100029, P.R. China - zhaodm@tea.ac.cn
Commission VI, WG VII/5
KEY WORDS: Soil moisture, Time series, ERS scatterometer, Floods, Droughts, Climate change
ABSTRACT:
Soil moisture is a governing parameter in many complex environmental processes from the disciplines of meteorology, hydrology
and agriculture. Since rainfall is partitioned into runoff and infiltration, the soil moisture content allows for direct information on
further infiltration capability and expected runoff behavior. Spatial and temporal soil moisture variability are thus important factors
to be included into predictive agricultural, hydrological and climate models. Furthermore, long term soil moisture pattern analyses
can support the derivation of regional trends. In this paper we present the results of analyses of the 15 year long, remote sensing
based soil moisture time series of TU Wien. Based on ERS scatterometer derived data soil moisture has been derived at a spatial
resolution of 50km, and a temporal resolution of 3-4 days globally since 1992. This time series is currently being extended and
reprocessed with 25km Metop Ascat derived data. We have processed the time series with respect to global anomaly derivation,
whereas an anomaly in the soil moisture dataset depicts “wetter than normal” or “drier than normal” conditions with respect to the
long term mean. Findings indicate that extreme events such as confirmed floods and droughts are clearly represented in the dataset.
Anomaly analyses in months prior to known extreme events indicate that the time series holds a strong potential for flood early
warning activities. Furthermore, long term trend derivation allows to depict regions, which have become significantly wetter or drier
over the course of the last 15 years. Trends investigated for Mongolia and Australia correlate with trends from in-situ station data.
We consider the TU Wien time series to have a high potential for further detailed global long term trend analyses.
1. INTRODUCTION
1.1 Remote Sensing of Soil Moisture
Soil moisture is a governing parameter in many complex
environmental processes from the disciplines of meteorology,
hydrology and agriculture. Since rainfall is partitioned into
runoff and infiltration, the soil moisture content allows for
direct information on further infiltration capability and expected
runoff behavior. Spatial and temporal soil moisture variability
are thus important factors to be included into predictive models
(Scipal et al. 2005, Wagner et al. 2007a, Koster et al. 1999,
Zhao et al. 2006, Pellarin et al. 2006).
In-situ soil moisture measurements are costly and work
intensive to perform and are thus only available in limited
regions of the world (Scipal 2005, Hollinger and Isard 1994).
Major in-situ soil moisture networks exist in China, Russia, the
Ukraine, and parts of the US (Scipal 2002, Wagner et al. 2007b,
Jackson et al. 1999, Robock et al. 2000). However, long term
temporal coverage and sampling intervals vary strongly. Due to
these reasons and due to the very sparse spatial representation
these stations do not enable to represent country-wide-,
continental- or even global soil moisture patterns. Thus, remote
sensing has come to play a major role in large scale soil
moisture assessment during the past two decades (Engman and
Chauhan 1995, Wagner et al. 1999, Wagner et al. 2007a,
Jackson 1993). Due to the cloud cover problem approaches
based on optical or thermal satellite data are strongly limited for
global applications. Therefore, especially microwave remote
sensing based on instruments such as the scatterometers
onboard ERS-1, ERS-2, or the AMSR radiometer onboard of
AQUA as well as the advanced scatterometer, ASCAT, onboard
the new satellite METOP, are employed for the derivation of
large scale soil moisture products (Wagner et al. 1999, de
Ridder 2000, Njoku et al. 2003).
The significance of global soil moisture products has been
presented by numerous authors. Many of them applied the ERS-
1/2 scatterometer derived soil moisture time series provided by
Vienna University of Technology (in following: TU Wien data
set); for example to improve rainfall simulation in eastern China
Zhao et al. 2006), to establish moisture-runoff relationships for
different catchments (Scheffler et al. 2003, Scipal et al. 2005),
or for data assimilation purposes. Furthermore, numerous
authors successfully validated the ERS derived TU Wien soil
Corresponding author. Claudia Kuenzer, DFD of DLR, Postfach 1116, 82230 Wessling, Germany, email: claudia.kuenzer@dlr.de,
fon: +49 (0)8153-28-3280