■■■■■■■H
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B7. Beijing 2008
3. RESEARCH DROUGHT MONITORING IN
CHONGQING WITH REMOTE SENSING
3.1 The Status of Drought Monitoring with Remote
Sensing
The agriculture drought indicators in China are generally
divided into three categories: precipitation index, soil moisture
indicator, water balance indicator. Soil moisture is one of the
important drought indicators. It is the major parameter in
climate, ecological, agricultural areas. Drought often happens in
large areas and lasts for a period of time, it is easy to cause
serious losses. With the development of remote sensing
technology, multiphase, multi-spectral of remote sensing data
may provide information related with surface soil moisture and.
It can be used to evaluate soil water conditions and monitor
agriculture drought.
NOAA/AVHRR (Kogan.F, 1997), Landsat TM were often used
for drought monitoring. Now, MODIS plays an important role
in agriculture drought monitoring. The following are the major
methods.
i) Thermal inertia approach. It usually applies to bare land or
low vegetation coverage area for remote sensing drought
monitoring.
ii) Vegetation supply water index. This method is particularly
suitable for vegetation coverage area or the abundance of
vegetation in better.
iii) Energy index model. It is more suitable for soil moisture
monitoring.
iv) Temperature vegetation drought index. This method has
been used in regional drought monitoring and has achieved very
good results. And so is water stress index (WDI).
3.2 Soil Moisture Index
Soil moisture is an important index in Chongqing agriculture
drought monitoring. There are 170 soil moisture (SI)
observation points in Chongqing city. SI is observed from
March to October on every 3rd and 8th. The data used in our
experiment was August 4 to 9 2006, which was supported by
the Chongqing Municipal Climate Center. Drought degrees are
as follows: Soil relative humidity ^ 30%: special drought; 30
to 40 percent: heavy drought; 40 to 50 percent: moderate
drought; 50 to 60 percent: Light drought; 60 to 90 percent:
suitable or basic suitable; > 90%: too wet.
3.2.1 The Necessity of Researching Soil Moisture Index
Since a MODIS station has been constructed in Chongqing city.
It is convenient to use MODIS as a remote sensing data
resource in agriculture drought monitoring in Chongqing. The
greatest advantage of remote sensing drought monitoring is its
real-time, fast, and objective. The greatest disadvantage is lack
of ground verification. So it is necessary to combine weather
observation, the ground truth observation and remote sensing
information together to fulfil accurate soil moisture monitoring.
3.2.2 Affecting Factors of Soil Moisture
The major factor of agricultural drought monitoring is the crop
water stress. So, the growth of crops and temperature (surface
temperature, including the canopy temperature) are important
indicators. The expression of soil moisture function can be
expressed as the following function:
SMI = f (NDVI, LST)
(3-1)
Vegetation index is used by the most researchers, and it is the
factor in operational work. Many scholars research a variety of
vegetation indexes. After analyzing and comparing the
vegetation indexes, the test with MODIS data from July to
August in Chongqing city shows that there is no saturation in
normalized difference vegetation index (NDVI), so we decide
to use NDVI as vegetation index.
3.3 Soil Moisture Retrieval Practice in Chongqing Summer
of 2006
Soil moisture retrieval formula is deduced by regression of
remote sensing data and ground observation data:
SMI = 0.856x NDVIx -X > — -0.0189
LST
(3-2)
Where LST is land surface temperature (K). NDVI is obtained
by reflectivity of MODIS band 1 and band 2.
NDVI =———
K+r.
(3-3)
Where r j, r 2 are reflectivity of MODIS band 1 and band 2.
Figure 3-1 is Chongqing NDVI map on August 7 th , 2006,
without removing cloud area. The cloud areas were removed in
soil moisture calculation.
The results are shown in Figure 3-2 and Figure 3-3:
Figure 3-2 is the retrieval soil moisture map. Each pixel
corresponds quantitative soil moisture. The region with value of
0 is cloud area and not involved in the calculation of moisture.
Figure 3-3 is the classification of soil moisture map basing on
Figure 3-2. Regions 1, 2, 3 and 4 denote special heavy drought,
heavy drought, moderate drought and slight drought. In our
study the soil moisture only referred to the moisture of 0-10cm
below the land surface.
i
A ?
423