Assessment of land use changes in Mukdaham and Nakhon Phanom provinces (NE Thailand) by means of Remote Sensing

Diplomarbeit von Nils Rennenberg

Abstract:

Tropical land use changes, predominantly the clearing of tropical rain and monsoon forests, have long been recognized as a trend with dramatic consequences. FAO estimated the global loss of rain forest area at 0.6 to 0.9 % per annum in 1993. Most of it can be attributed to the conversion of forest lands to agricultural areas. In Southeast Asia, two countries have suffered from this phenomenon more than any other nations: the Philippines and Thailand. Between 1961 and 1975, the forest reserves of Thailand have been reduced from 57 % to 37 % of the total area, while at the same time the area put to agricultural use has almost doubled. Only a small part of that process was due to organized resettlement programmes; to an overwhelming extent the deforestation has been performed 'illegally' by spontaneous activities of the rural population. The share of forests further declined to 28.9 % in 1998, at a current rate of -0.7 % per year, meaning that Thailand's forest cover has roughly halved since 1960.

The author of this thesis stayed as a visiting researcher at the Asian Institute of Technology (AIT), immediately north of Bangkok, for 6 months from November 2000 until April 2001. The AIT is a small international university for graduate studies with students and teachers from more than 40 countries. The core institution for this scientific cooperation was the Asian Center for Research on Remote Sensing (ACRoRS) of the AIT's School of Advanced Technologies (SAT). Field work in Northeast Thailand was carried out in two legs in February and April 2001.

As an underlying principle of this study, it was attempted to incorporate problems from the realms of social as well as physical geography, i.e. to maintain a balance between questions of applied geography (regional studies) and remote sensing. This also means that no special focus was put on advanced RS methodology such as the development of new image processing techniques; the study is rather based on a somewhat holistic approach, joining aspects from many different fields of science as diverse as geology, geomorphology, climatology, agriculture and agricultural economics, sociology, ethnology, politics, and spatial planning.

Accordingly, the basic objectives were:

- to give a geographical description/characterization of the Northeastern Region of Thailand in terms of its physical properties as well as its social and economical peculiarities.

- to try a quantification of forest losses since 1972 for a limited area of this region.

- to attempt a change detection, investigating the land cover (and possibly land use) changes within a timeframe of about 30 years.

- to name possible reasons for these changes.

- to find out if there are any connections between the development and socio-economic factors in NE Thailand (e.g. extension of crop area because of poverty/sinking incomes/population growth?).

- to tackle the question if the problem can be solved with the existing data and to make an assessment of the RS methodology that was applied.

- and to address the issue of future prospects and suggest a regional development concept to improve the living standards of the rural population.

As far as the limitations of the study are concerned, the biggest restrictions are associated with the remote sensing part, where the possible findings through the evaluation of satellite images are limited by the characteristics of the available data: their spatial, spectral and radiometric resolutions. Another difficulty arises from the fact that the older data stems from a different sensor type (Landsat MSS, as compared to Landsat TM), raising the question of comparability.

The contents of the study can be summarized as follows:

First, an introduction to the problem of tropical land use changes and an overview of the objectives, scope and limitation of the study are given, as well as some explanations about the terms „deforestation“ and „agro-colonization“ and the characteristics of these processes in Thailand; the theoretical part closes with some notes on remote sensing in general and the Landsat system in particular. Next, a comprehensive regional part introduces the reader to the Northeast of Thailand, starting with physical factors such as the geology, geomorphology, soils, climate, hydrology, vegetation, but also covering the human aspects to a great extent: administration, ethnicity, culture, society, economy, land utilization, and the living conditions of the rural population in this region. These observations then serve as a basis to describe the local characteristics of the study area, which are also treated in great detail, with the main focus on agriculture and the most important cultivated crops. A methodological part deals with data, materials and tools used for the remote sensing work, the different image processing steps, the findings of the two field works, and a digital elevation model (DEM) of the study area, generated from ancillary GIS data. In the results section, the land use patterns of the years 1972, 1989 and 2000 are compared with each other, and the main reasons for the changes are tried to be explained; a critical assessment of these results and the methodology used in the study is also given. Finally, it is attempted irradiate some aspects of development planning in Thailand, especially in the Mekong Region, and recommendations for its future sustainable development are given.

An appendix of 43 pages includes additional topographic and thematic maps, classification results, charts, tables, statistics (accuracy assessments etc.), a list of the 67 GPS waypoints attained during fieldwork, as well as selected photographs from the study area.

Table of Contents:

	List of figures	4
	List of tables	6
	List of abbreviations	7
	Remark on the quotation of Thai authors	7
	Remarks on the spelling of Thai place names	7
	Remark on Thai administrative units	9
	Acknowledgements	10
	References	147
	Internet references	150
	Map references	151
	Glossary of Thai terms and map designations	152
	Glossary of Thai measurements	154
	Appendix	155
	Photographs	184
1.	Introduction	11
1.1	Statement of problem: tropical land use changes	11
1.2	Objectives, scope and limitation of the study	12
1.3	Cooperation with AIT	13
2.	Theory/literature review	13
2.1	Deforestation and agro-colonization	13
2.2	Spontaneous versus planned colonization in Thailand	15
2.3	Remote sensing + the Landsat System	17
3.	Overview: Northeast Thailand	20
3.1	Location, administration, population	20
3.2	Geology and geomorphology	22
3.2.1	General outline	23
3.2.2	The Khorat Plateau	29
3.2.3	Soils	36
3.3	Climate and hydrology	37
3.4	Vegetation	46
3.5	Economy and land utilization	51
3.5.1	The importance of the rural space and agriculture in Thailand	51
3.5.2	Per capita income and living conditions of the rural population	52
3.5.3	General conditions for crop growing and overview of cultivated plants	54
3.5.4	Farm size, farm income and land tenure	57
4.	Study Area	59
4.1	Selection	59
4.1.1	JICA PLANET Project	59
4.1.2	Mukdahan and Nakhon Phanom Provinces	61
4.2	Location, size and population	62
4.3	Topography	67
4.4	Geology and soils	69
4.5	Climate	72
4.6	Land use/land cover	74
4.7	Agriculture	77
4.7.1	Rice	81
4.7.2	Sugarcane	83
4.7.3	Cassava	85
5.	Methodology	89
5.1	Data, materials and tools	89
5.2	Impressions from field work	90
5.2.1	First visit, February 2001	90
5.2.2	Second visit, April 2001	93
5.3	Image processing, part one (Thailand)	94
5.3.1	Data pre-processing	94
5.3.2	Rectification and geocoding	94
5.3.3	Mosaicking	94
5.3.4	Masking	94
5.3.5	Visual Interpretation	95
5.3.6	Unsupervised classification	97
5.3.7	Supervised classification	98
5.3.8	Accuracy assessment	105
5.4	Image processing, part two (Germany)	106
5.4.1	Resizing of year 1972 image	106
5.4.2	Data calibration test	106
5.4.3	Choice of area for case study and masking	107
5.4.4	Unsupervised classification	108
5.4.5	Band ratios and vegetation indices	108
5.4.6	Tasseled Cap Transformation	114
5.4.7	Supervised classification	116
5.4.8	Accuracy assessment	122
5.5	Elevation model	124
6.	Results	126
6.1	Results from remote sensing, part one (Thailand)	126
6.2	Results from remote sensing, part two (Germany)	126
6.3	Reasons for land use changes	129
6.4	Current situation of forests in Thailand	130
7.	Conclusion and outlook	131
7.1	Assessment of methodology	131
7.2	Development planning in Thailand	131
7.3	Regional development planning and future trends in the Mekong Region	133
7.4	Summary: problem structure and recommendations for the study area	40