GISdevelopment.net ---> Application ---> Agriculture

A GIS and Remote Sensing based approach to develop cold storage infrastructure for horticultural crops: A case study for potato crop in Bardhaman district, West Bengal

S. S. Ray, S. Dutta, N. Kundu and S. Panigrahy
Agro-Ecology and Management Division, ARG, Space Applications Centre (ISRO)
Ahmedabad - 380 053
1IWM&ED, Calcutta, West Bengal 
shibendu_ray@hotmail.com


Horticultural crops cover a large number of fruits, vegetables, flowers which are highly perishable in nature. Post harvest losses estimated to be in the range of 20-40 per cent. Hence, emphasis is given to develop post harvest infra-structure like cold storage, food processing, packaging, market outlets etc. during the current plan period . Most of the cold storages are concentrated in and around the consuming markets. Thus, very little facility exists to cater to the marginal farmer’s requirement during the harvesting season. Potato is one of the most important vegetable crop. Though 90 per cent cold storage facilities of the country is for potato crop and located in the potato growing regions, still it falls far below the requirement. The National Co-operative Development Corporation is trying to promote the setting up of such storage facility in the co-operative sector.. Thus, a scientific approach to evolve a methodology to locate sites for cold stores which would be optimally utilised by the growers is required. A pilot study was taken up to analyse the demand and supply situation and evolve an optimum plan to locate cold stores using satellite remote sensing (RS) data and Geographic Information System (GIS). The study was done for potato crop in Bardhman district of West Bengal, a leading potato growing area.

Materials and Methods

 Study Area
Bardhaman district is one of the major potato growing districts of West Bengal. The study was conducted for three police stations of Bardhaman district, namely, Jamalpur, Kalna and Memari which account for 82.9 and 83.3 per cents of potato area and production of the district respectively. There are 35 potato cold storages in the study area.

Data used
Satellite based RS data of IRS 1C WiFS (188 m resolution) was used for potato crop map generation. For getting the road network and settlement location high resolution RS data of IRS 1C LISS -III (23 m resolution) and Survey of India toposheets were used. The information about the cold storage locations, cold storage statistics (capacity, ownership, area of jurisdiction ) and potato crop statistics were collected from Department of Agricultural Marketing, West Bengal.

Crop Map Generation
The RS data was georeferenced and the boundary mask of Bardhaman district was overlaid. The extracted data was classified using a Maximum Likelihood classifier. Field information collected synchronous to the satellite pass was use for the supervised classification.

GIS Analysis
Spatial database was made for all the spatial data, like, road network, settlement locations, cold storage locations using Arc/Info GIS software. The database of the statistics of the cold storages was also linked to the spatial data. The raster image of the crop map, generated from satellite data, was converted to points coverage. Two steps of analysis was used for locating optimum cold storage sites. Those are buffering analysis and location-allocation analysis. In buffering analysis buffers were created around villages and already existing cold storage locations. The buffer sizes for the cold storages varied as per the capacity of the storage. Village buffer sizes were also dependent upon the village sizes. A large number (96) of probable cold storage sites were marked using the following criteria :
  1. the cold storage should be near a major road,
  2. it should be within the buffer of a settlement,
  3. it should not be within the buffer of the already existing cold storages.
Using the already existing cold-storages and the new probable sites location-allocation analysis was carried out to a) locate the required number of optimum new sites and b) allocate crop points to cold storage.

Results and Discussion
  1. Analysis of Data for Existing Cold Storage:
    A database of the information collected regarding all the 35 cold storages in the study area were analysed using Dbase. The analysis showed that the existing cold storages have a total capacity of 281.3 thousand tonnes which is remarkably low compared to the total production of potato (738.9 thousand tonnes) the area (Table 1). Assuming that ten per cent of potato will be used for local consumption , there is still a storage infrastructure deficit of 383.8 thousand tonnes. The average storage capacity of the existing cold storages is 6.74 thousand tonne. Hence there is a need of at least 57 new cold storages, of the average capacity, to store the additional potato production.

  2. Potato Crop Map Generation
    IRS WiFS data was digitally classified using a Maximum Likelihood classifier with the help of field information. The overall classification accuracy achieved was 92.0 per cent (Panigrahy and Chakraborty, 1998). The high accuracy achieved with the medium resolution (188 m) data was due to the fact that the potato in the study area is mostly grown in concentrated area, especially near rivers. The total estimated potato acreage of the study area was 28.7 thousand ha. A potato crop map was generated by masking all other land cover classes. The pixels in the map were converted to points in the Arc/Info coverage for being used as demand points in locating and allocating cold storages.

  3. Buffering Analysis
    Buffers are the zones around the geographic features. The buffers were generated around the settlements (big villages and townships) and cold storages. The buffer radius for each cold storage has been calculated based on its capacity. A buffer of 111 m was created for 1,000 tonne capacity of a cold storage. This buffer size was decided by taking the Bardhmann district average yield value of 257.06 quintals per ha. The settlement buffer was of the radius of 500 m except for three large townships (Memari, Kalna and Jamalpur) where it was 2.0 km. These buffers were used for marking, interactively, a large number of probable new cold storage sites (96 in number) as per the criteria set in the previous section.

  4. Location -Allocation Analysis
    Minimum distance criteria was used in the location-allocation analysis to optimally locate the new cold storage sites from the probable sites and allocate the crop points to the cold storage locations. Since the crop points are generally not connected by road a radial distance method has been used. The allocation showed that 5 out 35 existing cold storage are so injudiciously located that not a single crop point is allocated to them. Even when the analysis was repeated for only those 35 existing locations 4 did not get any allocation. The average distance to travel from crop point to cold storage reduced from 3.1 km to 1.6 km by incorporating the 57 new sites (Table 2). The maximum distance of travel also decreased from 5.6 to 3.6 km. While the average load on the existing storages were 210.7 thousand quintals it would be decreased to 79.0 thousand quintal on by incorporating these new locations.

    All the new suggested cold storages have been classified into three groups based on the number of demands allocated to them. The first 11 sites are of highest priority are those sites which are catering more than 200 crop points. There is an urgent need to create infrastructure for new cold storage in these sites. The 5 original cold storage sites which are not getting allocated any crop site can also be relocated to the new sites suggested.
Conclusions
 The study shows that remote sensing and GIS can be used to develop a scientific approach to draw an integrated plan for cold storage, food processing, packaging sites etc. for fruit and vegetable crops.

Acknowledgement
 The authors are grateful to Dr. R. R. Navalgund, Deputy Director, RESA and Shri J. S. Parihar, Group Director, ARG for constant encouragement. The help provided by Shri N. S. Mehta, Shri N.P.Bhatt, Shri Ananth Rao and Dr. Anupama Rastogi for GIS analysis and Dr. Manab Chakraborty for digital analysis is duly acknowledged.

References
Panigrahy, S. and Chakraborty, M, 1998. An integrated approach for potato crop intensification using temporal remote sensing data. ISPRS J. of Photogramm. & Rem. Sens. 53:54-60.

Table 1.Statistics of potato crop and cold storage in three police stations as compared to Bardhman district. 
Parameter Bardhmann Study Area
Acreage (`000 ha) 34.6 28.7
Production (`000 t) 886.6 738.9
Cold Storage Number 59 35
Cold Storage Capacity (`000 t) 355.1 281.3


Table 2. Improvement achieved in potato cold storage allocation 
Parameter Existing Cold storage Suggested Cold Storage
(Existing + New)
Total No. Of Cold Storage 35 92 (35+57)
Travel Distance (Crop site to Cold Storage), km
Maximum 5.6 3.6
Minimum 607 1172
Average 3.1 1.6
Capacity Required (‘000 q)*
Maximum 614.0 367.0
Minimum  0.91 0.91
Average 210.7 79.0

*Average capacity of existing cold storage is 67.4 thousand quintal

© GISdevelopment.net. All rights reserved.