Home Articles Integration of RS, GIS and AHP for Hanoi Peri-Urban Agriculture Planning

Integration of RS, GIS and AHP for Hanoi Peri-Urban Agriculture Planning

Rajesh B. THAPA, MSc.
Specialization: Geography
Remote Sensing and Geographic Information System.
Research Assistant
STAR Program
School of Advanced Technologies
Asian Institute of Technology, Thailand
Tel: +66-2-524-6406
Fax: +66-2-524-5597
Email: [email protected]

Frederic BORNE, PhD
Specialization: Image Analysis
Image Processing, Texture Analysis
3D Landscape Modeling, Linkage between plant growth modeling and RS/GIS applications
Assistant Professor
Centre de Coopération Internationale en Recherche Agronomique pour le Développement
Département AMIS, Amélioration des Méthodes pour l’Innovation Scientifique
Programme AMAP, Modélisation de l’Architecture des Plantes

Michiro, KUSANAGI, PhD
Michiro, KUSANAGI, PhD
Specialization: Aerospace System Engineering
Space System Engineering
Visiting Professor/Coordinator
STAR Program
School of Advanced Technologies
Asian Institute of Technology, Thailand
Tel: +66-2-524-5577
Fax: +66-2-524-5597
Email: [email protected]

Pham Van CU, PhD
Pham Van CU, PhD
Specialization: Geomorphology
GIS and Remote Sensing Professor/Director
Institute of Geology
National Center for Sciences and Technology, Vietnam
Tel: +84-4-83-51493
Fax: +84-4-83-59335
Email: [email protected]


The peri-urban areas are heterogeneous in social composition, which is characteristic for fast changing environments (Allen, 2001). Agriculture system in the peri-urban areas are a ‘result’ of migration from rural areas, absorption of former rural enterprises (Bentinck, 2000), urban farmers seeking expansion possibilities in the fringes. Both urban and peri-urban farming systems are highly diverse, and generalizations are difficult to make (Nunan, 2000). Part of the reason for the observed growth in urban and peri-urban agriculture is due to its adaptability and mobility compared with rural agriculture. The speed of urban growth and land use change might raise many problems such as inadequate infrastructure, population and employment pressure, overcrowding, slum occurred from low-income groups, fresh nutrient rich food insecurity and environmental degradation (natural as well as social). The challenge of supplying nutritionally adequate and safe food to city dwellers is substantial. Accomplishing these task under conditions of growth and congestion demands that policy-makers seize opportunities for integrating resource management and planning efforts, understanding potential linkages between rural and urban areas, and anticipating the changing needs of a country’s citizens – both rural and urban (Nugent and Drescher, 2000). Urban food supply (especially fresh nutrient-rich food) increment, employment and income generation, urban environment improvement, global food insecurity reduction and preserving the natural areas are the major contributions of peri-urban agriculture (Borne et al 2003). The sustainable peri-urban agriculture also depends upon market linkages. The market linkage promotes the spatial integration through economic interaction (Rondinelli, 1985). Since the market town is the main channel through which rural people obtain basic goods and services in return for their agricultural products, the impact of the coordination of marketing systems can have widespread effects and provide substantial benefits to the farmer. Therefore peri-urban agriculture needs proper management and development so that it can contribute to the economy, environment and society. Remote sensing is a powerful tool for collection and classification of spatial data whereas GIS is a powerful tool for management and analysis of spatial data (Nualchawee, 1996).

Study Area and objective
Hanoi province is located in Red River delta, the rice bowl of Vietnam after Mekong River delta. The province comprises of seven inner urban districts (Hoan Kiem, Ba Dinh, Hai Ba Trung, Dong Da, Tay Ho, Thanh Xuan and Cau Giay) called Hanoi city (capital of Vietnam) and five surrounding peri-urban districts (Tu Liem in the west, Thanh Tri in the south, Dong Anh and Soc Son in the north and Gia Lam in the east) with agricultural activities (Rossi et al 2002). Furthermore, the province is divided into 229 small administrative units (102 in urban and 127 in peri-urban) called commune. The whole province was selected as a study area. The province spread over about 928km2 area with flat (0.3% slope) topography from northwest to southeast. Because of sunny and tropical weather along with monsoons, the climate is favorable to agricultural development in Hanoi (Thapa, 2003). The population of Hanoi province was 2.81 millions with 3.2% growth rate where 53.56% population were lived in urban area and rest of the 46.44% lived in peri-urban area. The statistics shows the rate of urbanization in Vietnam has been increased from 15% in 1960 to 25% in 2001 and expected to be doubled (32%) in 2015. The economic growth of Hanoi has been rapidly developed at the rate of 11.6% per year during 1991-2000, which was bigger than overall country’s economic growth i.e. 7.5%. Crop, Livestock and Aquaculture are major agriculture system in Hanoi that contributed the 4.5% GDP (HSYB 2001). In order to gain the higher productivity and profitability, the competition between demand and supply of land for different activities are playing crucial role in land use morphology that challenges to Hanoi urban and peri-urban land use/agriculture planners and managers. There is a very essential element called expansion of market linkages in Hanoi have to be considered which is a primary force in commercializing agriculture, diversifying production and expanding the spatial system of exchange. Therefore, the timely updated land use information and major linkage information (physical, social and economical) between peri-urban and urban areas are needed for resources management to get optimum benefit in sustainable manner. An integrated technique of RS, GIS and AHP may be suitable in Hanoi peri-urban agriculture planning. The main objective of this study is to prepare a decision scenario for peri-urban agriculture planners.

Database and methodology
Five parameters, land use, soils, water resource, road and market were setup. A Baseline Survey of Vegetables Market of Hanoi–2003 conducted by SUSPER Project (a consortium with CIRAD, France, RIFAV, Vietnam and AVRDC, Vietnam), Landsat TM image acquired on November 23, 2001, a soil and a topographic map (1:10000) prepared by VTGEO (Vietnamese Institute for Geology) and a Field Survey on July 2003 were used as input of study. ENVI, ArcView and SPSS software were operated to compute the results.

The baseline data was geo-coded and prepared a vegetable market input linkage map as a representation of peri-urban agriculture and its contribution to urban. The land use map was prepared using Landsat TM 30meter resolution image within 90% of accuracy level. The soils map was classified based on Bardy (1974). Rivers and ponds/lakes were extracted from Landsat TM based land use map and used as water resource parameter. Considering different level of water flow motion in rivers and ponds/lakes, the water proximities at aerial distance of 1km, 2km and 3km form rivers and 0.3km, 0.6km and 0.9km from ponds/lakes were computed. Similarly, the aerial distances of 1km, 2km and 3km form road network and 10km, 20km and 30km form market zone were assigned while computing the road and market proximity to peri-urban agriculture areas. Each parameter was further converted into 30meter grid and scaled into Very Good, Good, Fair and Unarable for peri-urban agriculture. The Saaty’s (1990) Analytical Hierarchical Process (AHP) model was used to identify the priority of the parameters. This model evaluates the consistent weight of each parameter through pair-wise comparison (Canada et al, 1996). The priority index was computed within 0.09 consistency ratio based on primary information collected with the Hanoi subject experts. Each parameter multiplying with their weights were linearly added and scaled into 0 to 3 as Unarable(0), Fair (1), Good (2) and Very Good (3) for peri-urban agriculture. Commune level analysis was carried out in whole peri-urban area. For peri-urban agriculture planning, two maps were prepared at commune level where unarable land classification map presents the distribution status of land and the other map presents the suitability of land to peri-urban agriculture. Both maps can be used as guideline for agriculture planning. The urban districts were excluded while analyzing suitability of peri-urban agriculture.

Results and discussion

The vegetable market linkage (figure 4.1) shows that most of the vegetable products are served from peri-urban districts i.e. Dong Anh, Gia Lam, Tu Liem and Thanh Tri (table 4.1). Some provinces of Vietnam and China are also providing the vegetables products into the Hanoi urban. About nineteen types of vegetable products were found while surveying in market center in urban area. According to the properties of vegetable products supplied in urban, it was further aggregated into four groups as Fruits, Brassica, Leafy and Root types vegetables. The survey exhibited that overall 59% of marketed vegetables are produced in the province itself. Within the province, the fruits, brassica, leafy and root vegetable products contribute to the urban market centers for 44.39%, 65.11%, 67.25% and 59.72% respectively. The Dong Anh district, northern peri-urban district of the provinces, contributes 25.51% of total vegetable input. The Gia Lam, Tu Liem and Thanh Tri districts support about 14%, 9% and 8% of various vegetable products to the urban dwellers, respectively. Very nominal amount of vegetable products is also marketed from two urban districts i.e. Cau Giay and Tay Ho. However, about 15% vegetable is brought from Ha Tay provinces whereas 13% and 4% out of total come from Hung Yen and Vinch Phuc provinces respectively. Other provinces such as Bac Giang, Hai Phong, Ha Nam, Bac Ninh, Lam Dong, Son La and Ha Nam contribute in a small amount of vegetable in some seasons. There were some vegetable brought from China also, which accounts for 1.92% of the total. The results show that there is a significant linkage between peri-urban and urban and outer provinces. Even though, it presents only overview of the vegetable markets or vegetable input in urban areas but silently, in turn, the farmers get directly benefited in terms of economic perspectives from these activities.


Figure-4.1: Vegetable Input Linkage

Table 4.1: Type and Level of Vegetables Input to Hanoi Urban


Provinces Fruits % Brassica % Leafy % Root % Overall %
Hanoi: Districts 44.39 65.11 67.25 59.72 59.12
Dong Anh 27.40 33.92 11.26 29.44 25.51
Gia Lam 7.57 17.06 8.34 24.17 14.28
Tu Liem 5.86 6.86 22.92 2.22 9.47
Thanh Tri 2.46 6.80 21.59 2.50 8.34
Cau Giay 1.10 0.46 1.63 0.00 0.80
Tay Ho 0.00 0.00 1.50 1.39 0.72
Others provinces 52.21 30.61 32.75 40.28 38.96
Hung Yen 23.10 6.82 8.97 12.78 12.92
Vinh Phuc 3.92 5.73 4.67 3.33 4.41
Bac Giang 0.25 0.00 3.55 0.00 0.95
Ha Tay 20.76 9.83 10.63 19.72 15.24
Hai Phong 0.14 0.00 0.00 0.00 0.04
Ha Nam 0.21 0.00 0.00 1.11 0.33
Hai Duong 0.75 2.97 0.09 3.33 1.78
Bac Ninh 0.74 0.00 4.84 0.00 1.39
Lam Dong 1.99 5.21 0.00 0.00 1.80
Son La 0.33 0.00 0.00 0.00 0.08
Ha Nam 0.00 0.06 0.00 0.00 0.02
3.40 4.28 0.00 0.00 1.92

Saaty’s model called AHP (Analytical Hierarchical Process) was used to determine the priority weights of the selected parameters (land use, soils, water resource, road and market). The Hanoi experts’ judgment in pair-wise comparison between peri-urban agriculture parameters are identified their respective priority weights as soil (37), land use (31), road (16), water resources (10) and market (6).

Using the integrated result of AHP, RS and GIS techniques, the arable land suitability assessment map (figure 4.2) presents 502.18km2 (table 4.2) is unarable that covers the 54% land of the province. The unarable land category covers built-up, forest, unfertile soil, and water bodies. These areas cannot be used for agriculture purpose. From the peri-urban agriculture perspective, only 46% (426.23km2) land of the province is found as arable. While assessing suitable area for peri-urban agriculture in arable land, 32% (298.36km2), 12% (113.35km2) and 2% (14.52km2) area of the total land are found as Very Good, Good and Fair respectively. As compared to other peri-urban districts, the Soc Son and Dong Anh have bigger coverage of arable land. Within the arable land, the Tu Liem district, Gia Lam and Dong Anh have similar percentages of very good land. A big area (28%) of Soc Son district is found as good for peri-urban agriculture.

Figure-4.2: Suitable Land for peri-urban agriculture

Table 4.2: Land suitability assessment for peri-urban agriculture

District Unarable Fair Good Very Good Total
Units KM2 Percent KM2 Percent KM2 Percent KM2 Percent  
Dong Anh 90.31 48.62% 0.18 0.10% 15.28 8.23% 79.96 43.05% 185.73
Gia Lam 93.71 53.47% 0.02 0.01% 5.93 3.38% 75.60 43.13% 175.26
Soc Son 130.99 42.90% 14.26 4.67% 85.02 27.85% 75.03 24.57% 305.30
Thanh Tri 61.61 61.99% 0.04 0.04% 4.35 4.38% 33.39 33.59% 99.39
Tu Liem 42.45 53.34% 0.01 0.02% 2.76 3.46% 34.36   43.18% 79.58
Urban Area 83.11 99.96% 0.00 0.00% 0.00 0.00% 0.03 0.03% 83.14
Total Area 502.18 54.09% 14.52 1.56% 113.35 12.21% 298.36 32.14% 928.41

Unarable and arable land classification was performed within peri-urban districts at commune level. There are 127 communes in five peri-urban districts. The Gia Lam district has 35 communes where as Dong Anh, Soc Son, Thanh Tri and Tuliem have 24, 26, 26 and 16 communes respectively. The figure 4.3 shows that 15 communes (table 4.3) out of total 127 have above 75% areas of unarable land in each. There are 54 communes, which have area of unarable land between 50-75%. Similarly, 48 communes have 25-50% and 10 communes have less than 25% of unarable

Figure-4.3: Status of Unarable Land

Table 4.3: Commune level status of unarable land in peri-urban districts

District Area in percentage/Number of Communes
Dong Anh 75 Total
Dong Anh 0 14 10 0 24
Gia Lam 0 14 13 8 35
Soc Son 10 11 5 0 26
Thanh Tri 0 4 16 6 26
Tu Liem 0 5 10 1 16
Total 10 48 54 15 127

There is inverse relation between arable and unarable land – “bigger the unarable land, smaller the arable land”. The record shows the arable land in Hanoi is shrinking day by day due to urban enhancement. Each commune consists some pieces of land as arable and unarable land. Therefore, suitability scaling was performed in available arable land at commune level. The table 4.4 and figure 4.4 demonstrate 93 communes have very good arable land where other 24 and 10 communes have good and fair land for peri-urban agriculture, respectively. All communes in Tu Liem district are very good for peri-urban agriculture. The Gia Lam district, Than Tri and Dong Anh have 32, 23 and 18 communes out of 35, 26 and 24 as very good for peri-urban agriculture, respectively. About twenty kilometer (in aerial distance) far from the capital, 14 communes out of 26 of the Soc Son district have good land. Very few communes (4) are found as very good in Soc Son district as compared to other districts.

Figure-4.4: Suitability Assessment of Arable Land

Table 4.4: Commune level suitability assessment of arable land in peri-urban districts

District Area in percentage/Number of Communes
Fair Good Very Good Total
Dong Anh 1 5 18 24
Gia Lam 1 2 32 35
Soc Son 8 14 4 26
Thanh Tri 0 3 23 26
Tu Liem 0 0 16 16
Total 10 24 93 127

The integrated technique of RS, GIS and AHP for peri-urban agriculture planning has allocated the land at commune level with different level of suitability. The revealed results for peri-urban agriculture planning were derived geographically in respect of political boundary after prevailing intensive judgments. Therefore, the planners from Hanoi province may implement their plan in the scenario allocated land in priority basis at commune level. Every level of suitability can be utilized for different purposes for example; the unarable land if it has open space can be used for establishing the industrial estates or housing after checking their requirements. The very good category communes can be used for producing business agriculture such as fresh nutrient rich food production, cash crops and so on. Good category communes can be further assessed for specifi

agriculture practices or some urban gardening purpose. The fair land for peri-urban agriculture can be used for agriculture food processing zone or medium level industries. The planners and decision makers of the Hanoi province may follow such scenario adding other necessary decision parameters while allocating the land for peri-urban uses and future urban landscape then the future urban area will be developed as environmentally safe.

This study was conducted in the framework of SUSPER Project (Sustainable Peri-Urban Agriculture for South East Asia) FSP-Project 2000-56 funded by the Ministry of Foreign Affairs of France. The authors wish to thank CIRAD, VTGeo and all the Vietnamese partners for their contribution to the field trip and baseline survey.

Allen, A. (2001). ‘Environmental Planning and Management of the Periurban Interface’. Key note Paper for the Conference Rural-Urban Encounters: Managing the Environment of the Periurban interface, London 9-10 November 2001.

Bentinck J. (2000). ‘Delhi’s Urban Growth and the Stray-Cattle Controversy. In: Urban Agriculture Magazine’. No 2, Urban Livestock, October 2000, RUAF, Leusden The Netherlands. Borne, F., J. P. Satornkich, and S.M. Anwar (2003). ‘Plant Modeling for landscape Changes Visualization, Application to a Peri-Urban Agricultural Area’. PMA03 Conference, Beijing, 13-16 October 2003.

Brady, N.C. (1974). ‘The Nature and Properties of Soils’. McMillian Publishing Company, NewYork/London Canada, J. R., W.G. Sullivan and J.A. White (1996). ‘Capital Investment Analysis for Engineering and Management’. Prentice-Hall, New Jersey 07458.

HSYB (2001). ‘Hanoi Statistical Year Book’. Hanoi Statistical Office. Nualachawee, K., (1996). ‘Advanced Technologies in Geographic Information System’. Asian Institute of Technology, Thailand.

Nugent, R. and A. Drescher (2000). ‘Urban and Peri-Urban Agriculture (UPA) on the policy agenda: Virtual conference and information market’. A joint venture of the FAO Interdepartmental Working Group (IDWG) – Food for the Cities (FFC) and the Resource Center for Urban Agriculture and Forest (RUAF/ETC), [Downloaded: Jan 28, 2003].

Nunan F. (2000). ‘Livestock and livelihoods in Hubli-Dharwad, India. In: Urban Agriculture’. Magazine, no. 2, Urban Livestock, October 2000, RUAF, Leusden The Netherlands.

Rondinelli, D.A. (1985). ‘Applied Methods of Regional Analysis: the Spatial Dimension of Development Policy’. Westview Press, Boulder and London.

Rossi, G., P.V. Cu, F. Quertamp, and O. Chabert (2002). ‘Atlas Infographique de la Province de HANOI’. VTGEO-Institut de Géologie, CNST Vietnam / UMR CNRS-IRD 9937 “REGARDS”, France Saaty, T.L. (1990). ‘Multi Criteria Decision Making-The Analytic Hierarchy Process’. RWS Publication, Ellsowrth Avenue, USA.

Thapa, R.B (2003). ‘Spatial Decision Support Model for Sustainable Peri-urban Agriculture:Case Study of Hanoi Province, Vietnam’. MSc. Thesis. Asian Institute of Technology