TY - JOUR
T1 - Improved drought monitoring in the Greater Horn of Africa by combining meteorological and remote sensing based indicators
AU - Horion, Stéphanie Marie Anne F
AU - Kurnik, Blaz
AU - Barbosa, Paulo
AU - Vogt, Jürgen
PY - 2010/5/1
Y1 - 2010/5/1
N2 - Drought is a complex and insidious natural hazard. It is hence difficult
to detect in its early stages and to monitor its spatial evolution.
Defining drought is already a challenge and can be done differently by
meteorologists, hydrologists or socio-economists. In each one of these
research areas, various indicators were already set up to depict the
development of drought. However they are usually considering only one
aspect of the phenomenon. The development of integrated indicators could
help to detect faster/better the onset of drought, to monitor more
efficiently its evolution in time and space, and therefore to better
trigger timely and appropriate actions on the field. In this study,
meteorological and remote sensing based drought indicators were compared
over the Greater Horn of Africa in order to better understand: (i) how
they depict historical drought events ; (ii) if they could be combined
into an integrated drought indicator. The meteorological indicator
selected for our study is the well known Standardized Precipitation
Index, SPI. This statistical indicator is evaluating the lack or surplus
of precipitation during a given period of time as a function of the
long-term average precipitation and its distribution. Two remote
sensing based indicators were tested: the Normalized Difference Water
Index (NDWI) derived from SPOT-VEGETATION and the Global Vegetation
Index (VGI) derived form MERIS. The first index is sensitive to change
in leaf water content of vegetation canopies while the second is a proxy
of the amount and vigour of vegetation. For both indexes, anomalies were
estimated using available satellite archives. Cross-correlations
between remote sensing based anomalies and SPI were analysed for five
land covers (forest, shrubland, grassland, sparse grassland, cropland
and bare soil) over different regions in the Greater Horn of Africa. The
time window for the statistical analysis was set to the rainy season, as
it is the most critical period for vegetation growth. Moreover the
behaviour of those indicators was also investigated during major
historical droughts reported in the Emergency Database (EM-DAT) of the
Centre for Research on the Epidemiology of Disasters (CRED, Leuven
Belgium). Results of both analyses will be discussed during the
conference.
AB - Drought is a complex and insidious natural hazard. It is hence difficult
to detect in its early stages and to monitor its spatial evolution.
Defining drought is already a challenge and can be done differently by
meteorologists, hydrologists or socio-economists. In each one of these
research areas, various indicators were already set up to depict the
development of drought. However they are usually considering only one
aspect of the phenomenon. The development of integrated indicators could
help to detect faster/better the onset of drought, to monitor more
efficiently its evolution in time and space, and therefore to better
trigger timely and appropriate actions on the field. In this study,
meteorological and remote sensing based drought indicators were compared
over the Greater Horn of Africa in order to better understand: (i) how
they depict historical drought events ; (ii) if they could be combined
into an integrated drought indicator. The meteorological indicator
selected for our study is the well known Standardized Precipitation
Index, SPI. This statistical indicator is evaluating the lack or surplus
of precipitation during a given period of time as a function of the
long-term average precipitation and its distribution. Two remote
sensing based indicators were tested: the Normalized Difference Water
Index (NDWI) derived from SPOT-VEGETATION and the Global Vegetation
Index (VGI) derived form MERIS. The first index is sensitive to change
in leaf water content of vegetation canopies while the second is a proxy
of the amount and vigour of vegetation. For both indexes, anomalies were
estimated using available satellite archives. Cross-correlations
between remote sensing based anomalies and SPI were analysed for five
land covers (forest, shrubland, grassland, sparse grassland, cropland
and bare soil) over different regions in the Greater Horn of Africa. The
time window for the statistical analysis was set to the rainy season, as
it is the most critical period for vegetation growth. Moreover the
behaviour of those indicators was also investigated during major
historical droughts reported in the Emergency Database (EM-DAT) of the
Centre for Research on the Epidemiology of Disasters (CRED, Leuven
Belgium). Results of both analyses will be discussed during the
conference.
M3 - Journal article
VL - 12
SP - 11567
JO - EGU General Assembly 2010, held 2-7 May, 2010 in Vienna, Austria, p.11567
JF - EGU General Assembly 2010, held 2-7 May, 2010 in Vienna, Austria, p.11567
ER -
