AGRICULTURAL LAND INVESTIGATION AND CHANGE DETECTION
BY INCORPORATING GIS AND SATELLITE REMOTE SENSING;
CASE STUDIES IN SICHUAN PROVINCE, CHINA
Genong Yu, Maozhao Yan, Liangxiu Zen, Huamao Zhou
Chengdu Subcentre of Agricultural Remote Sensing
CHENGDU 610066, P. R. China
Qiaoling Zhang, Shirong Li, and Haiqi Liu
Ministry of Agriculture
BEIJING 100026, P. R. China
Commision VII, Working Group 5
KEY WORDS: Geographical Information System (GIS), Agricultural Land, Remote Sensing, Environmental Change
ABSTRACT
Repeat agricultural land information requirements prompt the study of the potential application of remote sensing techniques
and geographical information system (GIS). A procedure for agricultural land investigation and change monitoring by
incorporating these techniques was summed up through the studies carried out recent years in Sichuan province, China, which
was proved feasible and operational. This paper also discussed thoroughly two case studies adopting the procedure. In
addition, some further research aspects were pointed out which should be in consideration to make the sort of procedures
more operational and more efficient.
1. INTRODUCTION
The needs for reliable, precise and up-to-date information
on agricultural land and its change are growing, since there
exist heavy demands from government ministries such as the
Ministry of Agriculture, the National Bureau of Land
Management, and diversities of regional and local
government authorities. This is also true even for some non-
departmental organizations, such as the Commission for
Construction. The present, annually available information is
mainly sourced from multipurpose statistical data, which do
not meet the specific requirements on agricultural land
information often because of inadequacy of type of data,
time and relative precision.
For satisfying these needs, series of projects have been
carried out since 1986, based on diversities of original data
and some specific requirements of information, such as
those on peat resource (Yu et al 1992), desertization, land
use, soil erosion, and forest. This paper describes a
feasible, operational procedure for agricultural land mapping
and change detection on a large scale by using Landsat
Thematic Mapper (TM) image or other high resolution
satellite images. Two case studies are presented to show
the carrying-out and the results of the approach.
2. OPERATIONAL PROCEDURE
In correspondence with the constant requirements of timely
information, series of projects have been carried out
nationwide. The following summarized the basic
requirements of these projects.
a. The information achieved on agricultural land and its
change should be representative on a large scale, to be
useful for national or provincial decision making.
b. Both time and accuracy is important concerns. It is not
acceptable that the investigation is highly accurate but too
time-consuming, or instant but unreliable results.
c. The information should be obtained within a reasonable
cost. Normally, there is no sufficient budget to support the
carrying-out of aerial photography or the acquirement of
digital satellite image in these projects.
d. Change detection is a long-range and repeat task.
Investigation will be repeated on the same area and for the
same purpose. Information will be renewed annually.
Therefore, the procedure used by the project is mainly
based on the visual interpretation of Landsat TM image in
print with preprocessing and geometric correction,
completed in the light of the available historical data and the
routine ground survey. This may be the cheapest and
quickest way to fulfil these tasks in a developing country.
See Fig. 1.
ay Non — remotely — sensed data
x Y
opographical tic [Ground
Landsat TM map map | survey | data
Y
Visual viron. | Geo.
interpretation | |Pase mapi factors | unit
Correction
MAP coefficient
i
Digitizing &
editing
i
GIS & OVERLAYS &
MAPS INFORMATION DYNAMIC
Figure 1. Flow Chart of the Agricultural
Land Change Monitoring
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996
