Home Articles The integration of Remote Sensing system and GIS for forest landuse planning

The integration of Remote Sensing system and GIS for forest landuse planning

ACRS 1995

Poster Session 1

The Integration of Remote Sensing System and GIS for Forest Landuse Planning

Dr. Suwit Ongsomwang
Forest Resources Assessment Division
Forest Research Office, Royal Forest Department
Bangkok 10900. Thailand


Abstract

In general, the applications of remote sensing to forestry in Thailand were mostly based on visual interpretation and the thematic data that were produced from the remote sensing data were still separated. The accuracy map acquired from the interpretation of aerial photographs at 1: 15,000 scale for forest and non-forest area in the past cannot be used as a base map for the interpretation of satellite imageries of scale 1: 250,000 at present. In fact, it leads to the inconsistency of the extraction data from Landsat imageries and also increases the time and cost for the ground truth survey. Actually, the integration of the thematic data with remote sensing data is limited because of using visual interpretation. method. The main object of this study is to develop reliable method for integration of remote sensing system and the GIS for forest land use planning.

The method was developed by using the multistage remote sensing concept and the integration of remote sensing system and the GIS. In this study, Landsat images were used as the first stage for existing forest area classification. In the secondary stage, aerial photographs at the scale of 1: 15,000 were used for forest land use classification. In the third “stage, ground verifications were conducted for collecting and generating the nieces information for forest land use planning. In addition, the geographic information databases that were designed for handling and processing spatial data of remote sensing and ancillary data were established under the exchangeable format of the image processing system and the GIS.

The main forest land types in the forest land use plan of the Ngao Demonstration Forest, Ngao District, Lampang Province are (I) non-forest area with no limiting conditions, (2) non-forest area with limiting conditions, (3) mining area, (4) community forestry area, (5) productive reforestation area, (6) productive forest area, (7) protective. reforestation area and (8) protective forest area.

The results from this study shows that the integration of remote sensing and the GIS appears to have potential for providing the necessary information for forest land planning, particularly in classification of existing forest land uses and their status. In addition, forest land use planning which was conducted by using the GIS shows the new methodology to the manager and planner for making a forest land use plan.


I. Introduction

The application of remote sensing data in forestry in Thailand was mostly based on visual interpretation and the thematic data that were produced from the remote sensing data were still separated. The accuracy map that would be interpreted for forest and non-forest area in the past by using aerial photographs of scale 1: 15,000, for example, cannot be used as a basic map for the intetrpretation of satellite imageries of scale 1: 250,000 at present. In fact, it leads to the inconsistency of the extraction data from Landsat imageries and also increases the time and cost for the ground truth survey. .Actually, the integration of the thematic data with remote sensing data is limited because of the visual intetrpretation. In addition an efficient geographic handling and processing system that transforms the thematic data into usable information does not exist.

At present the major tool for the geographic handling and processing system of the thematic data is the Geographic Information System (GIS). The GIS is a set of tools for collecting, storing, retrieving at will, transforming, and displaying spatial data from the real world for a particular purpose (Burrough 1986). The GIS evolved as a means of assembling and analyzing diverse data pertaining to specific geographic areas, using the spatial locations of the data as the basis for the information system (Shelton and Estes 1979). Kalensky (1992) stated that remote sensing and the GIS are closely related. They are complementary information technologies which are both needed and operationally used for natural resources mapping and inventories, natural disaster assessment and risk forecasting and for environmental monitoring. If they operated separately in institutes for management of natural resources, it would be harmful. It would result in costly duplication of various activities, training and some equipment and negatively affect the efficiency of project implementation. Goodenough (1988) also suggested that for the integration of remote sensing and the GIS, the remote sensing data must use the digital image analysis system. The main objects of this study are to develop reliable method for integration of remote sensing system and GIS for forest land use planning.


2. Study Area

Ngao Demonstration Forest area was chosen as the study area, The Ngao Demonstration Forest area extends over the headwaters of Ngao river in the north-west of Lampang province, Thailand down to its confluence with the Yom river on the Phrae province boundary, the boundaries being those of Ngao district, Most of the Ngao Demonstration Forest area that is in Ngao district, Lampang province lies between 18° and 21,5 north, 99° and 45 east The total area is “”; about 1,751,59 sq, kIn,

3. Methodology

The basic concept for the development study method is employed from the “multi II remote sensing concept which was , enunciated by Colwell (1975). Multistage remote sensing concept is used in this study. This concept encompassed the ;i: variety of remote sensing platforms and sensors that can be used in various combination, to acquire remote sensing data and also considered the processing, analysis and utilization methodologies available to practitioners in the field of remote sensing. In the multistage remote sensing approach, satellite data may be analyzed in conjunction with high and low altitude data, and ground observation. Each successive data source might provide more detailed information about
geographical areas. Information extracted at any lower level of observation may then be extrapolated to higher levels of observation (LiIlesand and Kiefer 1979). In addition, the geographic information databases that were designed for handling and processing spatial data of remote sensing and ancillary data were established under the exchangeable format of the image processing system and the GIS. McFarland (1982) stated that the establishment and maintenance of geographic databases in computer-readable form has great .potential for supplying resource management information.

To fulfill this methodology, a scheme of pract1cal work on this study 1S shown m Figure 1



Figure 1 Schematic flowchart of the main tsks in this study

ACRS 1995

Poster Session 1

The Integration of Remote Sensing System and GIS for Forest Landuse Planning


4. Results and Discussions

4. 1 Forest land utilization classification
Forest land utilization is classified by the normalized indexing operation of the four GIS images which represent the natural conditions and regulated conditions for forest land utilization in the study area. The areas and percentage of . four input GIS images with their assessment and ranking for the normalized indexing operation are summarized in . Table 1, The results of the normalized indexing operation are shown in Table 2. The results show that a relatively ., important value from 1 to 8 implies the suitability of forest land utilization from lowest to highest, respectively. In addition, the results also show the various alternatives in the classification. In this study, the determination of the three .forest land utilization classes is a personal decision with the main objective of increasing the protective forest areas in the study area. The forest land utilization classification of the study area consists of unsuitability for forest area suitability for productive forest area and suitability for protective forest area. They cover an area of 39,745.08 ha, ..57,236.22 ha, and 78,178.14 ha, respectively. The results show that the proportion of the three forest land utilizations in the study area has been upgraded in comparison with the input GIS images. r .

Table 1 The area and percentage of the four input GIS images used in the forest utilization classification.

Watershed classification as forest land use capability by natural conditions
No. Class Ranking Area in ha Percent
1 Unsuitability for forest area 1 55,00836 31.41
2 Suitability for productive forest area 5 68,430.06 39.07
3 Suitability for protective forest area 10 51,721.02 29.52
3 units Total 175,159.44 100.00
Soil classification as forest land use suitability by natural conditions
No. Class Ranking Area in ha Percent
1 Unsuitability for forest area 1 47,408.85 27.07
2 Suitability for forest area 5 127, 750.59 72.93
2 units Total 175,159.44 100.00
National forest reserve as productive forest areas by regulated conditions
No. Class Ranking Area in ha Percent
1 Unsuitability for forest area 1 12,176.19 6.95
2 Suitability for productive forest area 5 162,983.25 93.05
2 units Total 175,159.44 100.00
Forest and national parks as protective forest areas by regulated conditions
No. Class Ranking Area in ha Percent
1 Unsuitability for forest area 1 104,308.29 59.55
2 Suitability for protective forest area 10 70,851.15 40.45
2 units Total 175,159.44 100.00

Table 2 The combination of four input GIS images by the normalized indexing operation.

No. Value Relative suitability for forest utilization Area in ha Percent
1 1 Very unsuitable for forest area 11,429.46 6.53
2 2 Unsuitability for forest area 28,315.62 16.17
3 3 Suitability for productive 27,154.62 15.50
4 4 Very suitable for productive forest 30,081.60 17.17
5 5 Suitability for protective forest 25,273.98 14.43
6 6 Moderate suitability for protective forest 16,396.02 9.36
7 7 Suitability for protective forest 157.41 0.09
8 8 Very suitable for protective forest 36,350.73 20.75
8 units Total 175,159.44 100.00

4. 2 Combination of forest land utilization classification with a regulated forest utilization

In this study, the Temporary Cultivation Rights (STK) areas and mining concession areas GIS images, which represent the existing forest utilization by regulations in the study area, are combined with the forest land utilization classification ~ by the maximum overlaying operation and the results are shown in Table 3. The results show that forest areas in the forest land utilization classification decreased from 135,414.36 ha to 133,404.84 ha after the combination existing forest utilization by regulations. 4. 3 Evaluation of the forest land utilization classification with the existing forest land use in 1989. In this study, the existing forest land use in 1989 GIS image, which represents the social and economic factors affecting
the natural resources in the study area, is combined with the forest land utilization classification with a regulated forest
utilization GIS images by the normalize? indexing operation. The evaluation of forest land utilization and existing forest land use 1989 are separately summarized as follows:

Table 3 The combination of forest land utilization and existing forest utilization by regulations.

No. Forest land utilization and existing forest utilization Area in ha Percent
1 Unsuitability. for forest area 37,017.72 21.13
2 Suitability for production forest area 55,920.06 31.93
3 Suitability for protection forest area 77,484.78 44.24
4 Temporary Cultivation Rights (STK) areas 4,372.02 2.50
5 Mining concession areas 364.86 0.21
5 units Total 175,159.44 100.00

I. Unsuitability for forest area. The results show that existing forest areas which include the Mixed Deciduous forest, Dipterocarp Deciduous forest and productive reforestation areas overlap with the unsuitability for forest area, In fact,
the existing forest areas cover an area of 15,991,92 ha or about 9,13 percent of the total study area, In the forest land use plan, the existing natural forest areas are assigned as the community forestry areas, while the existing productive reforestation areas are assigned as productive forest areas, In contrast, the existing non-forest area in the unsuitability for forest areas are assigned as the non-forest areas with no limiting conditions in the forest land use plan.

II. Suitability for productive forest area. The results show that non-forest areas which consist of the settlement and built-, up areas, agricultural areas, old clearing areas and mining areas overlap with the suitability for productive forest area. Areas of the existing non-forest are 7,648.56 ha. The main component of the existing non-forest areas is the old if clearing area which covers an area of7 ,348.14 ha or about 4.20 percent of the total study area. These areas are assigned as productive reforestation areas in the forest land use plan. In the meantime, the existing settlement and built up areas and agricultural areas are here assigned as non-forest areas with limiting conditions. In contrast, the existing forest areas in the suitability for productive forest area are assigned as productive forest areas in the forest land use plan.

III. Suitability for protective forest area. The same as with the suitability for productive forest area, the results show that the non-forest areas which consist of the agricultural areas and the old clearing. Areas exist .in the suitability for protective forest area. The main component of the non-forest areas IS the old cleaning area which covers an area of ‘ 27,047.25 ha or about 15.44 percent of the total study area. These areas are assigned as protective reforestation in the forest land use plan. In the meantime, the existing settlement and agricultural areas are here assigned as non-forest ~ areas with limiting conditions.

IV. Temporary Cultivation Rights (STK) area. The results show that existing forest areas overlap with the Temporary Cultivation Rights (STK) areas. This fact is contradicted with the principle of the Temporary Cultivation Rights projects which operate in the deteriorated forest land. In this study, all of the Temporary Cultivation Rights (STK) areas are assigned as non .forest areas with limiting conditions in the forest land use plan.

V. Mining concession area. The results shows that most of the active mining areas in 1989 overlap with the mining
concession areas. The status of mining concession areas is unchanged in the forest land use plan.

ACRS 1995

Poster Session 1

The Integration of Remote Sensing System and GIS for Forest Landuse Planning


4. 4 Forest land use plan

In this study, a forest land use plan is a possibility plan based on the evaluation of the relationship between the forest land utilization classification including the existing regulated forest utilization and the existing forest land use in 1989. In practice, the forest land use plan map is firstly produced by the recoding operation of the output from the unnormalized indexing operations and then a suggestion of the forest land utilization as the main content of the forest land use plan is given. The forest land utilization with suggestion in the forest land use plan are described as follows:

I. Non-forest area with no limiting conditions. The non-forest area with no limiting conditions includes the Temporary Cultivation Rights (STK) areas and the existing non-forest areas in 1989 which overlap with the unsuitability for forest , area of the forest land utilization classification. It covers an area of 21,025.80 ha or about 12.00 percent of the total study area. In the execution of the plan, the existing old clearing areas from 1989 which were the main component in these areas should be promoted for the intensive agricultural utilization. In addition, the clarification of land rights and land reform for these areas are here suggested.

II. Non-forest area with limiting conditions. The non-forest area with limiting conditions includes the Temporary Cultivation Rights (STK) areas and the existing non foes area in .1989 which overlap with the suitability for’ productive and protective forest areas of the forest uuhzatlon classification. It covers an area of 4,888.17 ha or about2.79 percent of the total study area. In the execution of the plan, the soil conservation protection measures and the clarification of land rights for these areas are required.

III. Mining area. The mining area is mining concession areas in the study area. It covers an area of 364.86 ha or about .
0.21 percent of the total study area.

IV Community forestry area. The community forestry area is. the. Existing .natural forest areas in 1989 which overlap With the. Unsuitability for forest area of the forest land .utilization classifications. In. this study ‘. the assignment of community forestry area is intended to decrease the activity of illegal tree cutting which occurs m the natural forest areas. In addition, this area can be considered as the buffer zone between the intensive productive or protective forest area. The community forestry area covers an area of 13,836.60 ha or about 7.90 percent of the total study area. This area should be divided into small areas for each Tambon (sub-district) in the execution of the plan. The management and utilization of the community forestry area are determined by the committee of the Tambon with the suggestion of
the Royal Forest Department. Basically, trees in the existing Mixed Deciduous forest are suitable for timber utilization , with the selection systems, while trees in the existing Dipterocarp Deciduous forest are suitable for firewood production ! with the coppice systems.

V. Productive reforestation area. The productive reforestation area is the existing old clearing areas in 1989 which overlap with the suitability for productive forest area of the forest. land utilization classification. It covers an area of , 7,348.14 ha or about 4.20 percent of the total study area. The traditional planting system with Teak species is suggested fu in the execution of the plan. ,

VI. Productive forest area. The productive forest area consists of the existing natural forest and productive reforestation. ,
areas in 1989 which overlaps with the suitability for productive forest area of the forest land utilization classification. In addition, the .existing productive reforestation areas .in I :89 which overlap the unsuitability for forest area of the forest land utilization classification is also Included in this area. The productive forest area covers an area of 50,426.82 ha or about 28.79 percent of the total study area. This area has a main function of sustainable productive forest. In the . execution of the plan, intensive forest inventory is required for the logging operation. In addition, the traditional selection system with a 30-years cutting cycle and a selection felling ratio of 60 percent for trees over girth limit (commercial trees) should be revised as the actual forest recovery forest situation. According to the forest inventory data in 1991 the number commercial trees in the natural forest is relatively low in all forest types and stratum. Furthermore, the termination of forest logging in the Mae Ngao demonstration forest management unit has proved to be the failure of the traditional selection system in the past. The clear cutting system in the small forest areas with a low gradient is here suggested. In practice, the traditional planting system is immediately operated after forest logging.

VII. Protective reforestation area. The protective reforestation area is the existing old clearing areas in 1989 which overlap with the suitability for protective forest area of the forest land utilization classification. The protective r reforestation area covers an area of 27,047.25 ha or about 15.44 percent of the total study area. In the execution of the plan the traditional planting system with growing tree species is suggested for the old clearing areas without small trees from the natural succession. In contrast, enrichment planting system are suggested for the old clearing areas with some trees from the natural succession. In practice, trees from natural succession in reforestation sites should be preserved and forest fire control is required.

VIII. Protective forest area. The protective forest area includes the existing natural forest areas and the protective
reforestation in 1989 which overlap with the suitability for protective forest area of the forest land utilization classification. The protective forest area covers an area of 50,221.80 ha or about 28.67 percent of the total study area. In fact, this area has the main function of environment protection. In execution of the plan, tree felling is prohibited in this forest area.


5. Conclusion

In conclusion the integration of remote sensing and the GIS therefore appears to have potential for providing the necessary information for forest land use planting, in particular the existing forest areas and their status. In addition, forest land use planning can be conducted easily and effectively by using the GIS as a tool. However, the successful of the integration of remote sensing and the GIS for forest land use planning depends on the reliability of the existing information which are sources for the GIS, for example, reforestation map.


6. Reference

  • Burroughs, P. A. 1986. Principles of Geographic Information Systems for Land Resources Assessment. Oxford University
    Press, Oxford. 193 p.
  • Colwell, R. N. 1975. Introduction. In Manual of Remote Sensing, First Edition (Editor R. G. Reeves) American Society
    of
    Photogrammetry. Virginia: I: 25.
  • Good enough, D. G. 1988. Thematic Mapper and Spot Integration with a Geographic Information System. Photogrammetric Engineering and Remote Sensing, 54(2):167 -176.
  • Kalensky. Z. D. 1992. FAO Remote Sensing Activities in Environmental Monitoring and Forest Cover Assessment in Developing Countries. Invited paper. ISPRS Commission VII, Session on Tropical Forest and Land Use Monitoring in Seventeenth Congress of the International Society for Photogrammetry and Remote Sensing, Washington, D. C., USA, 2-14 August 1992. 9 p.
  • Lillesand , T. M. and R. W. Kiefer. 1979. Remote Sensing and Image Interpretation. John Wiley & Sons. New York. 612p
  • McFarland, W. D. 1982. Geographic Databases for Natural Resource. In Remote Sensing for Resource Management. (Edited by C. I. Johannsen and I. L. Sanders). Soil Conservation Society of America. Iowa: 41-50.
  • Shelton, R. L. and I. E. Estes, 1979. Integration of Remote Sensing and Geographic Information System. In
    Proceedings .
    of the Thirteenth International Symposium on Remote Sensing of Environment, Volume I: 675-692.