965
2008
A DUST-STORM PROCESS DYNAMIC MONITORING WITH MULTI-TEMPORAL
MODIS DATA
DI Mei, LU Xiushan, SUN Lin, WANG Ping
Geomatics College, ShanDong University of Science and Technology,ShanDong Qingdao 266510,China
dimei666@ 163.com, sunlin6@ 126.com,pingwsd@ 163 .com
KEY WORDS: Dust Storm; Dynamic Monitoring; Multi-Temporal Data; Terra/Aqua MODIS
ABSTRACT:
In recent years, the frequency of dust storm is rising. Dust storm can impact the climate changes and cause great damages to people.
It is very necessary to monitor the disasters accurately. In this paper, a dust storm process occurred in April, 2006 has been monitored.
According to the reflect and absorb characteristic of dust storm, the scope of dust storm is extracted successfully;and the intensity
grades are estimated, then the moving trace is monitored by multi-temporal data. Result shows greet agreement with the
meteorological department monitoring. Conclusion can be gotten: the dust storm process dynamic monitoring with multi-temporal
data has great application prospect.
1. INTRODUCTION
Dust storm, a general name of sand storm and dust storm, is the
weather condition which means strong wind draws much sand
and dust near the ground surface into the sky, making air very
foul, causing visibility less than 1km .Dust storm is a complex
process influenced by the interaction of earth-atmosphere
system, mainly caused by high wind speed, bared soil and dry
air condition etc. It often takes place in arid and semi-arid areas.
In recent years, the frequency of dust storm is rising. Dust storm
can impact the climate changes and cause great damage to
people. It is very necessary to monitor the disasters timely and
accurately.
At present, the ground based measurement method and remote
sensing technology are two main methods for monitoring dust
storm (Li Haiping, Xiong Liya,Zhuang Dafang, 2003). When
dust storm happens, the environmental condition is often very
bad. The sites for measuring dust storm are usually laid sparsely.
As a result, because of low temporal and spatial resolution,
traditional ground based measurement method cannot meet the
requirements of dust storm monitoring and forecasting very well.
The technology of satellite remote sensing has many advantages,
such as: wide coverage, continuous in the space and monitoring
natural disasters quickly, so it can act as an important role in the
dust storm monitoring. Remote sensing can monitor the scope
of dust storm, its intensity grade and its moving trace.
Lots of researchers in China and abroad have done numerous
experiments on monitoring dust storm by the technology of
satellite remote sensing. Luo Jingning (Luo Jingning,Fan
Yida,Shi Peijun,2003) has constructed a comparable dust storm
intensity index to solve this problem: because of different
satellite platforms ,monitoring time and areas, a comparable
monitored result cannot be given by using multi-source remote
sensing data; Guo Ni (Guo Ni,Liang Yun,2006) has constructed
two dust indexes to discriminate dust storm processes happened
in 2002-2005; Li Qing (Li Qing,Wang Qiao,Wang Wenjie,2006)
has monitored a strong dust storm process occurred in the
northwest of China in 2003 by MODIS data; Liu Sanchao (Liu
Sanchao,Liu Qinhuo,Gao Maofang,2006) has extracted dust
storm areas by combining Terra and Aqua images at daytime
and nighttime;John J. Qu(John J. Qu, Menas Kafatos, 2006) has
proposed a normalized difference dust index(NDDI) to identify
dust and cloud; J. K. Roskovensky (J. K. Roskovensky ,K. N.
Liou, 2005)has combined short-wave reflectance ratio tests with
long-wave brightness temperature differences to produce
individual parameters to detect cirrus and dust. However, many
researchers only focused on accurately identifying whether a
pixel is contaminated by dust storm and then limit the areas of
dust storm distribution. In these years, more and more satellites
have been sent to the space, it has become possible to make
certain the moving trace of the dust storm using the
multi-temporal satellite images obtained from different satellite
sensors. In this paper, the scope of dust storm is extracted
successfully and then the moving trace and intensity grades are
monitored by multi-temporal data. The results can help related
departments grasp the moving trends of dust storm in
macroscopic view and provide a basis for them to make
decisions.
2. DATA
In this paper, the data used to monitor dust storm disaster is
MODIS data. MODIS sensors, boarded on both Terra and Aqua
satellites, have 36 channels. Its spectrum range is 0.4pm~
14.385pm,covering from visible to infrared. At least 4 MODIS
images can be obtained every day. The timeliness of data is
increasing, so dust storm can be monitored dynamically. In this
paper, 4 MODIS images are chose to monitor a dust storm
process happened in northwest of China on April 16-17, 2006
(see table 1).