GISdevelopment.net ---> AARS ---> ACRS 1989 ---> Poster Session

Use of high resolution Remote Sensing data for land use investigation in Vietnam

Pham Trung Luong, Truong Thi Hoa Binh
Remote sensing Application Division Center of Geography and Natural
Resources NCSR of Vietnam


Introduction
The use of remote sensing data as Landsat 1, 2, 3 for study and investigation of Landuse in Vietnam has been carried out since 1980 and has given good results that permit to overcome and disadvantages of traditional method. However, in many cases, due to Laduse characteristics of developing country likes Vietnam and different Landuse traditions of different geographic regions, a more detailed investigation and classification on higher level are demanded for: - statistic work; study of destruction of agricultural land and natural forest under natural and human actions detection of changes of Landuse, crops and irrigation; - investigation of reclaiming waste land and forest replantation. In those cases, Landsat 1,2,3 images are not satisfied and high-resolution data likes Landsat 4 TM and SPOT images are required. They permit to revolutionize Landuse investigation using remote sensing technology. These will be discussed in this paper with results of some experiments using available data.

High Resolution Remote Sensing data analysis for landuse investigation
Landsat 4 TM images are not available in Vietnam that is why only SPOT images are used as high-resolution data for more detailed land use investigation of different geographic regions of the country.

Spectral resolution of SPOT are three spectral bands 9visble and near infrared portions of the spectrum) with 20m of ground resolution and a broader spectral band (Panchromatic B and W) with 10m of ground resolution.

In order to study landuse patterns based on remote sensing techniques, three bands in visible and near IR part of the spectrum are mainly used. The first one is the green (First maximum of the reflectance) the second one is the red (minimum of the reflectance due to

Chlorophyll absorption) and the third is near IR where the reflectance of active vegetation for Landsat MSS: Band 4,5 and 7 for Landsat TM: bands 2,3, and 4 and for SPOT bands 1,2 and 3. Usually above bands combinations have been shown by color composite on which the first band is printed in blue, the second - in green and third - in red.

Available SPOT data here does not ever entire Vietnam. There are only some POST image of regions (Mekong river delta, Northeastern part of Red river delta and central plateau (Scheme 1).

Used data for this study are the followings.
  1. For Mekong river delta:

    • SPOT image (B and W) dated 7/3/1986, 20/6/1986 and 13/1/1987 at scale 1/250,000
    • SPOT FC dated 7/3/1986, 27/1/1987, 31/2/1987, 8/2/1987 and 1/3/1987.

  2. For Central Plateau (TAY NGUYEN):

    • SPOT (B and W) dated 20/4/1987 at scale 1/250,000
    • SPOT FCC dated 20/4/1987 at scale 1/250,000
    • Toposheets at scales 1/250,000. 1/100.000. 1/50,000
Used analysis method is the visual interpretation on electroptical equipments as zoom transferscope. Additive Color Viewer AC-90B acts…. In order to verify the pre-field classification, some field checks have been done also. Color code blue, green and red for SPOT bands 1,2 and 3 paddy fields of Mekong river delta appear in various shades of red tone; for Landsat MSS bands 4,5 and 7 - in yellow, orange and various shades of red tone. Due to stand density, moisture content and species, vegetation cover appears in tonal different (Yellow, orange and red). Indicates that SPOT images give more information that allows classify more detail the crops and soils (Tab.1)

Information ability of Landsat MSS and SPOT images for land use investigation of mekong river delta
Table. 1
Landuse pattern Landsat MSS SPOT
Not able to detect Able to detect Able to detect (clearly) Not able to detect Abel to detect Able to detect (clearly)
1. Agricultural land
* Wetland rice
- after harvest
- young
- mature
*Upland crop
*Perennial trees
- Orchard
. Coconut
. Others
- Commercial
2. Forestry land
* Mangrove
- Dense
- Secondary
- Plantation
* Vegetation on Acid
sulphate soil
3. Settlement
* Urban
* Village
4. Land for special use
* Salt field
* Rea for shrimp nurse
* Rush
5 Others
* Sand, Sand dune
* Barren land and scrubs
* Marsh
* Road
*Channel





X






X




















X
X








X





X



X
X





X
X
X
X





X
X


X
X



X






X








X


















X


















































X
X
X
X
X

X
X
X








X
X
X


X
X

X


X
X
X
X
X

Another region is Daklak province, which has high potential of forest, coffee, tea and rubber plantation. People in this province, is minor, nations mainly. Their tradition sis shifting cultivation. They cut the forest, grow the corps and after harvest they abandon the used lands, which became waste, and go to other place and being to cut the forest and etc…. again. due to above mentioned cause, landuse changes very quick. This change can be detected on SPOT image.

Interpretation result of SPOT images is shown in table 2. On SPOT FCC (of bands 1, 2 and 3) landuse patterns are seen in various of yellow orange and red tone and we can classify landuse patterns in detail and we can study and monitor deforestation and reforestation.

Information Ability of Landsat MSS and spot images for Landuse investigation of centeral plateau
Table 2
Landuse pattern Landsat MSS SPOT
Not able to detect Able to detect Able to detect (clearly) Not able to detect Abel to detect Able to detect (clearly)
1. Agricultural land
* Wetland rice
- after harvest
- young
- mature
*Upland
* Dry rice
* Upland crop
* Coffee plantation
* Tea garden
* Perennial Trees
- Orchards
- Commercial
2. Forestry land
* Evergreen forest
- closed (density 78//)
- dense (density 40-80/)
- Open (Density (40/)
*bamboo
* Mixed
Plantation (Pine)
3. Settlement
* Urban
* Village
4. Land for special use
* Grasse
* Dam, Water reserver
5 Others
* Barren land and scrubs
* Road
*Irrigation system





X





















X







X

X


X

X








X

X



X

X









X
X




X




X


X



X

X




X







X

X

X
X









































X

X
































X
X
X



X
X
X

X
X


X

X
X

X
X
X


X
X


X




X
X

Conclusion
Using high-resolution Remote Sensing data ahs more advantages than MSS data. That kind of a data can satisfy the practical demands of land use investigation in Vietnam at present.

For landuse investigation and its planning on sub district level (at scales 1:100,000 and 1:50,000) use of high resolution remote sensing data has economic effect.

Use of the high resolution remote sensing data combined with other data likes aerial photos or ground data permit to investigate landuse upto scale 1/25,000.

References
  • Mikhiro Ioka and Masato Koda. Performance of Landsat 5 TM data in land covers classification. Int. Journal Remote Sensing 1986 Vol. 7 No. 12.
  • C.P.LO. Lisses and remote sensing Longman scientific technical 1986
  • Thomas M, Lissessand Remote Sensing and Image Interpretation.
  • Pham Trung Luong, Vo Duc Tuyen Nguyen Ba Nhuan, Truong Thi Hoa Binh and Tran Minh Y.
  • Ability of application of remote sensing technis for compiling landuse map on the large scale in Vietnam.