Land Suitability Analysis for Rice Cultivation Based on Multi-Criteria Decision Approach through GIS

Transcription

1 Land Suitability Analysis for Rice Cultivation Based on Multi-Criteria Decision Approach through GIS Sailesh Samanta #1, Babita Pal #2 and Dilip Kumar Pal #3 # 1, 3 Department of Surveying and Land Studies, PNG University of Technology, Private Mail Bag, Lae- 411, Morobe Province, Papua New Guinea #1 rsgis.sailesh@gmail.com #3 dkpal2000@yahoo.com # 2 PNG University of Technology campus, Lae- 411, Morobe Province, Papua New Guinea #2 mundey2@gmail.com 12 Abstract- Geographic Information System (GIS) has demonstrated itself as a very powerful tool in agricultural research and natural resource management. This study proposes an empirical methodology for analyzing and mapping of land suitability using the GIS techniques. The study area happens to be the Morobe province in the Papua New Guinea. This paper examines multi-criteria decision approach to determine land suitability for rice cultivation based on different variables, like topography (slope and aspect of the land), physical (texture, water holding capacity and depth) and chemical (ph, nitrogen, potassium, phosphorus) soil properties, climate (temperature and rainfall) and land accessibility that are mandatory inputs to land suitability model. These parameters are obtained from Shuttle Radar Topography Mission (SRTM) data, soil data base of PNG, monthly and annual temperature and rainfall data, respectively. ArcGIS 10 and Erdas 11 model builder/maker are used to construct the index model for rice land suitability analysis. The province of Morobe has been classified into five categories of rice suitability. The result indicates that only four percent (4%) land can be demarcated as high and twenty one percent (21%) as medium high suitability categories in the study area and the spatial expanse of all the five categories within the province are mapped and displayed. Keywords: Geographic Information System, soil, land suitability, rice, vegetation cover 1. Introduction Crop-land suitability analysis is a prerequisite to achieving optimum utilization of the available land resources for sustainable agricultural production (Perveen et al., 2007). One of the most burning needs in Papua New Guinea (PNG) is to improve agricultural land management and to impart suitable cropping patterns in order to increase the agricultural production with efficient use of land resources. This is paramount towards augmenting agriculture s contribution of country s GDP, which is much below International Journal of Science & Emerging Technologies IJSET, E-ISSN: Copyright ExcelingTech, Pub, UK ( the inherent potential at present. The bulk of the Papua New Guinea s population rely on subsistence farming of non-cereal-grain crops, such as taro, yam, banana, and sweet potato, and/or exploitation of sago palm, which constitute the staple food for majority of Papua New Guineans. One of the most striking human ecological characteristics in this country is seen in different population densities in association with the environments where they have lived and the major foods which they have grown and eaten. Traditional food-production systems in Papua New Guinea can be broadly classified into four categories: two in the high-altitude zone and two in the lowaltitude zone. In the high-altitude zone, sweet potatodependent agriculture has been dominant in the central Highlands and taro-dependent agriculture in the Highlands fringe. In the low-altitude zone, cultivation of various crops, such as taro, yam, and banana has been common in the eastern part of the main land and the islands region, whereas exploitation of sago has been prevailing in the western part of the main land. In Papua New Guinea rice is one of the major foods, but the production is very low. In particular the rice production of PNG is hardly keeping pace currently with the rapidly rising demands due to its very low production and steeply rising demand from enlightened people who have started preferring a swap from the traditional nongrain staple food. The situation perpetuates PNG s dependence of import from south-east Asian countries. Instead of her turning into a net exporter of rice by facilitating rice agriculture through identification of suitable rice growing areas, PNG continues to be heavily dependent on rice imports. Hence this study has been envisaged with a view to determining physical land suitability for rice crop using a multi-criteria decision and GIS approach. Finally the utilisation of the potentially rice-suitable lands vis-à-vis the current vegetation cover / land use on the same tract is explored. The aim in integrating Multi-Criteria Evaluation with Geographical Information Systems (GIS) is to provide more flexible and superior mechanism to the decision makers in order to evaluate the effective factors. This research provides information at local level that could

2 13 be used by farmers to select cropping patterns in accordance with suitability. Market access map 2010 PNG University of Technology 2. Study Area and Data Used This work attempts to develop index model to determine suitable land for rice cultivation in Morobe province of Papua New Guinea. The study area is bounded within ' to 148 E longitude and 5 to 8 S latitude. Different types of data sets are used for rice land suitability analysis. Various physical and chemical variables are used for this purpose, they are topography (slope and aspect of the land), physical (texture, water holding capacity and depth) and chemical (ph, nitrogen, potassium, phosphorus) soil properties, climate (temperature and rainfall) and land accessibility. Optical bands with standard false color combination of Landsat-7, TM satellite images are used to find out the canopy cover classes in the study area. One of the most widely used digital elevation model (DEM) data sources is the elevation information provided by the shuttle radar topography mission (SRTM) (Coltelli et al. 1996), but as with most other DEM sources, the SRTM data requires significant levels of pre-processing to ensure that there are no spurious artifacts in the data that would cause problems in later analysis such as pits, spikes and patches of no data (Dowding et al. 2004). In the case of the SRTM data, these patches of no data are filled, preferably with auxiliary sources of DEM data, like topographical maps. Both data sets are used for this study. Different statistical data are obtained from Geobook 2009 census of Papua New Guinea to carry out this research work. Climate Research Unit (CRU) of East Anglia University, UK has developed a database of monthly gridded (0.5 x 0.5 ) climate observations globally using various techniques (Mitchell and Jones, 2005). This data set contains nine climate variables, namely-precipitation, daily mean temperature, monthly average daily minimum temperature, monthly average daily maximum temperature, diurnal temperature range, vapour pressure, cloud cover, wet day frequency and frost day frequency. Temperature and rainfall variable have been taken for this study. All other details of the variables, data spans along with the sources were given in the table 1. Table 1 - Different data used for land suitability analysis Variables Year Source Soil data and 2009 Geobook maps Topographical maps 1960 University of Texas Libraries, Austin SRTM data 2003 ftp://e0srp01u.ecs.nasa.gov Landsat-7, TM 2002 University of Maryland Institute Temperature and rainfall Methodology Spatial multi-criteria decision-making approach is a process where geographical data is combined and transformed into a decision. It involves input data, the decision maker s preference and manipulation of both information using specified decision rules. In this spatial multi-criteria decision-making approach, the input data is geographical data. Topography, physical and chemical properties of soil, climate and land accessibility are used for the land suitability analysis for rice cultivation Topographic Database The original SRTM DEM is used to produce slope and aspect for the study area. Processing is made on a void by void basis. In cases when a higher resolution auxiliary DEM is available, point coverage is produced of the elevation values at the centre of each cell of the auxiliary DEM within void areas. When no high resolution auxiliary DEM is available, the 30 second SRTM30 DEM is used as an auxiliary for large voids (Hutchinson (1988; 1989). On the other hand, DEM is also generated from topographical maps. Erdas-11 is used to rectify topographical maps of the study area using the Universal Transverse Mercator (UTM) projection and WGS 84 datum with a RMS error of Using the 3-D analysis process the digital elevation model is generated from contours and spot heights, which have been digitized from topographical maps. Prepared data sets in this process are compared with the SRTM data set to find out the accuracy of both data sets. The Slope tool calculates the maximum rate of change between each cell and its neighbors. Every cell in the output raster has a slope value. The lower slope value indicates a flatter terrain and higher the slope value steeper is the terrain. Aspect identifies the steepest down slope direction from each cell to its neighbors. It can be thought of as slope direction or the compass direction a hill faces. Aspect is measured clockwise in degrees from 0, due north, to 360, again due north, coming full circle. The value of each cell in an aspect dataset indicates the direction the cell's slope faces. Flat areas having no down slope direction are given a value of - 1. ArcGIS is used to generate slope and aspect map of the study area using the DEM data. Flat fields having smooth surface are better for rice cultivation as it facilitates even and equal distribution of water Soil Database Physical and chemical data base of soil are prepared from soil region maps of Geobook. In the United States, twelve soil texture classifications are defined by the United State Department of Agriculture (USDA), namely sand, loamy sand, sandy loam, silt

3 14 loam, loam, sandy clay loam, silty clay loam, clay loam, clay, sandy clay, silty clay and Silt. Clay, silt clay, silt clay loam, textures of soil are best for paddy/rice crop. The effective soil depth is defined as the thickness of soil above a layer restricting root growth (e.g. consolidated rock or cemented materials, such as gravel). Most annual crops have a rooting depth of about 50cm, while for tree crops the rooting system can reach beyond 150cm. However, most crops produce good yields in soils with an effective soil depth of about 100cm and this value has been used as an upper limit. The available water holding major soil elements N, P, and K are generally correlated with plant growth and subject to rapid changes following forest clearing and cropping. For each case, we converted them into 3 groups according to their characteristics for rice cultivation Climate Data Temperature and rainfall are two climatic factors which has a favorable and in some cases unfavorable influence on the development, growth and yield of rice. Rice being a tropical and sub-tropical crop is normally grown at a fairly high temperature high rainfall regime, ranging from 20 to 40 C and Direct access to market Indirect access to market Table 2 - Travel time and distance from village to access market Mode of access Travel time from Distance from village (Km) village By walk By motor car By boat Minor market <1hour <4 <40 <22 Major market < 2 hours <8 <80 <44 Capital and central market <3 hours <12 <120 <66 Mobile communication <1hour <4 <40 <22 Airstrip transportation < 2 hours <8 <80 <44 Major road/transport < 2 hours <8 <80 <44 Major wharf < 2 hours <4 <40 <22 Source: Samanta et al., 2011 Table 3 - Topography suitability rating for rice Slope (%) Rating Aspect direction Rating Less than 15 1 Flat, N, NE & NW less than 15 percent slope 1 15 to 30 2 All directions more or less than 15 percent slope 0 More than 30 3 Table 4 - USDA soil texture suitability rating for rice Textural classes Description (USDA) Rating Sands More than 85% sand, 0 to 10% clay and 0 to 15% silt 3 Loamy Sands 70 to 91% sand, 0 to 15% clay and 0 to 30% silt 3 Sandy Loams More than 43% sand, less than 7% clay and less than 50% silt 3 Sandy Clay Loam More than 45% sand, 20 to 35% clay, less than 28% silt 3 Sandy Clay 45% or more sand and 35% or more clay 2 Loam Less than 52% sand, 7 to 27% clay, 28 to 50% silt 2 Silt Loam Less than 12% clay and 50 to 80% silt 2 Clay Loam 20 to 45% sand and 27 to 40% clay 2 Silt Less than 12% clay and 80% or more silt 2 Silty Clay Loam Less than 20% sand and 27 to 40% clay 1 Silty Clay 40% or more clay and 40% or more silt 1 Clay Less than 45% sand, 40% or more clay and less than 40% silt 1 capacity of soil is the amount of water held in the soil between tensions corresponding to field capacity and permanent wilting point. The ph of the soil is defined as the negative logarithm of the hydrogen-ion concentration of the soil solution. Slightly acid soils having a ph value of 6 to 7 are better for paddy cultivation. However, it has been found to be grown in a wide range of ph varying from 4 to 8. The three 1250mm to 2000mm of annual rainfall. Different raster data sets are generated by the interpolation process using these climate variables of Climate Research Unit. 3.4 Market Access Data Base We consider market access as one indirect variable for determination of suitable land for cultivation as one input in the multi-criteria decision making

4 15 model. Existing data base (Samanta et al., 2011) for Morobe province is used for this study. Village point, informal market, formal market, provincial capital market, telecommunication, airstrip transportation, major road transport and major wharf data set are used for this analysis (Table 2). Proximity analysis is performed using village points to generate market access zone using all market access parameters. Overlay analysis of each market access zone with village point layer is processed to find out whether the villages are accessible to any kind of market or not. Then village points within each market access area are coded as 1 and village point out of market access area 2. Statistical analysis (SUM) is performed of all coded village according to different mode of market access. Final village point layer is generated with total market access code and classified into three (3) market accessible ranks, like 7 to 9 as good market access (1), 10 to 11 as moderate market access (2) and 12 to 14 as poor market access (3). 3.5 Suitability Rating According to the degree of favorable environment for rice crop, simple statistical weighting/ratings were used for all the variables leading to multi-criteria decision support approach. We contrived three rating systems, like 1 as suitable, 2 as moderately suitable and 3 as unsuitable for all variables (Figure 1). All the details about weighting/rating are given in table 3 through 8. Table 5 - Soil depth and available soil water holding suitability rating for rice Available water holding capacity Rating Soil depth description Rating Very high (>15cm) 1 Not stony/rocky 1 High (10 to15) 1 Slightly stony/rocky 1 Moderate (5 to 10) 2 Moderately stony/rocky 2 Low (3 to 5) 3 Very stony/rocky 3 Very low (<3) 3 Extremely stony/rocky 3 Table 6 - Soil ph, nitrogen (N), phosphorus (P) and potassium (K) suitability rating for rice Soil ph Rating Nitrogen (%) Rating Phosphorus Rating Potassium Rating 6.0 to High (> 0.5) 1 High (> 20 ppm) 1 High (> 20 ppm) to Moderate 2 Moderate 2 Moderate 2 (0.2 to 0.5) (10 to 20 ppm) (10-20 ppm) 5.0 to Low (<0.2) 3 Low (<10 ppm) 3 Low (<10 ppm) 3 < > Table 7 - Climate (temperature and rainfall) suitability rating for rice Temperature (degree C) Rating Rainfall (mm) Rating 10 to to to to to 40 3 More than Table 8 - Market access suitability rating for rice Market access code Accessibility Rating 7 Very High 1 8 High 1 9 Medium high 1 10 Medium 2 11 Medium low 2 12 Low 3 13 Very low 3 14 Inaccessible 3

5 16 Figure 1 - Suitability rating map of slope (a), aspect (b), soil texture (c), water holding capacity (d), depth (e), soil ph (f), soil N (g), soil P (h), soil K (i), temperature (j), rainfall (k) and access (l).

6 Land Suitability Analysis Erdas-11 and ArcGIS-10 software were used to prepare topography layers (slope and aspect of the land), soil physico-chemical layers (texture, water holding capacity, depth, ph, nitrogen, potassium, and phosphorus), incumbent major climate feature layers (temperature and rainfall) and land access rating layers (Figure 1). We devised the relevant index model (Chart 1) in the model maker using the multi-criteria 4. Results and Discussion Figure 1 shows all twelve variables, namely topography (slope and aspect of the land), physical (texture, water holding capacity and depth) and chemical (ph, nitrogen, potassium, phosphorus) soil properties, climate (temperature and rainfall) and land access, which are all used in the index model according to the suitability rating criteria. ArcGIS platform is used to prepare these rating maps after Chart 1 - Methodological flow chart for land suitability analysis decision-making approach. All those twelve variables in figure 1 were used as inputs in the index model. In the first step we produced the topographic suitability using slope and aspect layer for the area, followed by physical soil suitability using texture, water holding capacity and depth layer; chemical soil suitability using soil ph, nitrogen, potassium and phosphorus layer; climate suitability using temperature and rainfall layer; and finally land access suitability composites. In the next step all suitable categories were stored in the temporary output memory file. Finally we used them as temporary memory input for rice crop land suitability analysis. reclassifying all attributes for each variable. Statistics for each rating class are calculated for all twelve variables as shown in the table 9. The rating and sum functions are used in the index model (Chart 1) to produce the final output map of suitable rice crop land for Morobe province (Figure 2).

7 18 Figure 2 - Land suitability zones for rice crop cultivation, based on multi-criteria decision-making approach using slope and aspect of land, texture, water holding capacity, depth, ph, available nitrogen, potassium, phosphorus of soil, temperature, rainfall and land access of Morobe province, Papua New Guinea by Erdas-11 and ArcGIS-10.

8 19 Table 9 - Spatial distribution statistics of each suitable rating zone for variables for rice cultivation Variables Categories % of Area according to the suitability rating Rating-1 Rating-2 Rating-3 Topography Slope Aspect Soil physical properties Soil texture Water holding capacity Soil depth Soil chemical properties Soil ph Nitrogen (N ) Phosphorus (P) Potassium (K) Climate Temperature Rainfall Access Land access Figure 3 Vegetation cover of Morobe province, Papua New Guinea for, based on hybrid maximum-normalized Differential Vegetation Index (NDVI) and minimum-red compositing technique.

9 20 Canopy cover data set is generated (Figure 3) from satellite images using a hybrid maximum-normalized Differential Vegetation Index (NDVI) and minimumred compositing technique. The present vegetation cover map and the suitability map for rice crop are overlaid to identify differences between the present vegetation cover and the potential suitable land for rice. A cross table between the map of suitable areas and the vegetation cover map is obtained. In this way, we obtained useful information concerning the spatial distribution of different suitability levels. This phase allowed us to fine-tune our results, because the resultant layer provided the information about how the rice can be cultivated across the various land suitability zones. The overlay analysis is performed between the rice suitability map and vegetation cover map of the study area for better understanding of their relationship. The index model predicts the inverse relationship between the percent of vegetation cover and rice land suitability. All the medium, medium-low and low suitable lands are laid under dense vegetation cover (more than 50%), where as high and mediumhigh suitable lands are cited in the open vegetation cover (less than 50%) in the study area (Table 10). Table 10 - Relationship between suitable land for rice cultivation and vegetation cover Land suitability for % of land rice area % of vegetatio n cover High to 50 Medium-high 21.3 Medium to 75 Medium-low Conclusions Low to 81 Rice crop suitability analysis is carried out for entire Morobe province, Papua New Guinea. Spatial multicriteria decision-making approach is used with twelve geographical data sets, as input in the model, namely slope and aspect, soil texture, water holding capacity, depth, ph, nitrogen, potassium, phosphorus, temperature, rainfall and land access. Markham valley region in the middle, north east part of the study area and major parts of Lae district comes under high to medium high rice suitable zone, where vegetation cover is very negligible. We received better result after cross checking our modeled output with the existing maps of Geobook data set of the study area. In the future study we can attempt small scale mapping (national level) as well as large scale (district level) mapping on similar theme for Papua New Guinea with addition and further refined parameters. laboratory facility to carry out the research work. The authors are also grateful to the all the academic staff of GIS section and National Agriculture Research Institute for their valuable comments and suggestions. References [1] Coltelli, M., Fornaro, G., Franceschetti, G., Lanari, R., Migiaccio, M., Moreira, J. R., Papathanassaou, K. P., Puglisi, G., Riccio, D., and Schwabisch, M., SIR-C/X-SAR multifrequency multipass interferometry: A new tool for geological interpretation, Journal of Geophysical Research, 101, [2] CRU (Climate Research Unit), average temperature and rainfall data over 100 years, TS 2.1, [3] Dowding, S., Kuuskivi, T., and LI, X., Void fill of SRTM elevation data Principles, Processes and Performance, In: Images to Decisions: Remote Sensing Foundations for GIS Applications", ASPRS, Fall Conference, September 12-16, Kansas City, MO, USA, [4] Hutchinson, M., Calculation of hydrologically sound digital elevation models, Third International Symposium on Spatial Data Handling, Columbus, Ohio, International Geographical Union, [5] Hutchinson, M. (1989), A new procedure for gridding elevation and stream line data with automatic removal of spurious pits, Journal of Hydrology, 106, pp , [6] Mitchell, T. D. and Jones, P. D., Aan improved method of construction a database of monthly climate observation and associated highresolution grid", International Journal of Climatoogy, 25, pp , [7] Perveen, F., Ryota, N., Imtiaz, U., Hossain, K. M. D., Crop land suitability analysis using a multicriteria evaluation and GIS approach, 5th International Symposium on Digital Earth, The University of California, Berkeley, USA, pp. 1-8, [8] Samanta, S., Pal, D. K., Antonio, W., and Pal, B., Spatial Modeling and Interpolation to Establish Market Accessibility using ArcGIS v-10.0, 4th Science and Technology Conference 2011, (27 th June to 1 st July), Vudal Campus, East New Britain, Papua New Guinea, pp. 8. [9] SRTM (Shuttle Radar Topography Mission), elevation data on a near-global scale, ftp://e0srp01u.ecs.nasa.gov. [10] USDA (United State Department of Agriculture), Soil texture classification, Acknowledgements Author expresses sincere gratitude to Papua New Guinea University of Technology & Department of Surveying and Land studies for providing GIS