498
Application of Remote Sensing and GIS for Crop Inventory – Crop Discrimination, Acreage and Yield estimation – A case study Saharanpur district, Uttar Pradesh State (India) MANSOUR D.(1), SAHA S.K.(2) & PATEL N.R.(1) (1)
Center of Space Techniques, 01 Avenue de la Palestine BP13 31200 Arzew. Oran, Alegria,
[email protected] Indian Institute of Remote Sensing, Dept. of Space, Govt. of India, 4, Kalidas Road, Dehradun - 248 001 (India),
[email protected]
(2)
Abstract. In This paper we present an approach based on the application of remote sensing techniques for crop discrimination, acreage and yield estimation, optical data was used for crop identification in rabi and kharif crop seasons. Remote sensing now become a very interesting tool in agriculture, so in this present work, we tried to highlight the contribution of satellite imagery for crop identification, acreage and yield estimation. This study has been carried out for Saharanpur district of Uttar Pradesh State for crop acreage estimation and yield distribution, during Rabi and kharif seasons of the year 2005-2006. Two types of data have been used in this study – IRS 1C- LISS III for October 15, 2005 and IRS 1CLISS III for February 12, 2006. In this approach, the district administrative boundary of Saharanpur district was overlaid over the remote sensing image to extract the image Saharanpur district. Then crops area were identified and estimated by following supervised maximum likelihood digital classification. Statistic results of images classification of February and October show accuracies (91.41% and 97% respectively for October and February). Otherwise, the distribution of yield map was achieved from the relationship NDVI and yield measured in the field. This map has shown that yield around 32 and 40 q/ ha represents 39% of the total area of Saharanpur district. Keywords: remote sensing, digital classification, crop inventory, acreage, yields. 1. Introduction. The Indian government gives great importance to agriculture; both planners and policy makers are still interested in this sector and consequently reliable agricultural statistics over time. Policy makers still need in time of agricultural statistics for the most important crop in India namely wheat, Sugarcane, rice and other crops (Maize, cotton...). The spatial of crops distribution, acreage and yield prediction are the interesting information in the sustainable management of agriculture. [8] Remote Sensing (RS) using space-borne sensors is a tool, par excellence, for obtaining repetitive (with a range from minutes to days) and synoptic (with local to regional coverage) observations on spectral behavior of crops as well as their growing environment, i.e., soil and atmosphere. Use of this data could be made for a number of applications such as crop inventory, crop production forecasts, drought and flood damage assessment, range and irrigated land monitoring and management. 2. The objectives of this study. The present study attempts to fulfill the following objectives: Discrimination and acreage estimation of [9] kharif and Rabi crop seasons using digital processing techniques of satellite data IRS – 1C LISS III. Yield prediction of [9]rabi crops using relationship between NDVI and yield 3. The Study area. Saharanpur district is bounded on the two sides i.e. in north by Dehradun and in west by Haridwar districts of Uttaranchal, on east by Ambala and Karnal district of Haryana state and on south by Muzaffarnagar district of Uttar Pradesh. The total area of district is 3689 km2 with a density of 785 persons per square km.
Fig. 1 Location of the study area
Colloque International des Utilisateurs de SIG, Taza GIS-Days, 23-24 Mai 2012 Recueil des Résumés
499
4. The methodology followed. The first step was to prepare the land use/land cover map, from the supervised classification of the multispectral data of IRS-1C-LISS3 accomplished by the ground truth collection.
Flow diagram of methodology of crop inventory using LISS III data (kharif, rabi)
The second step was to prepare the yield map, from the NDVI and the mask applying on the classified image of rabi crop season to obtain wheat crop only. The relationship between the wheat NDVI and the yield shown a better correlation, because major part of cultivate area was under wheat crop ([7] Yield (Q/ha) = 60.84*NDVI – 9.895, (Adj. R2 = 0.521, SEE = 7.142 N = 44)) 60 50 yield (Q/ha)
40
Graphe.1: linear regression shows the correlation between yield and NDVI.
◊ Field data (N=44,Wheat yield)
30 20 10 0 0.2
0.4
0.6
0.8
1
NDVI
5.
Results and discussion. a. Crop inventory and others land use/land cover. Area under the land use / land cover was derived from classified image of kharif season (fig.2) and the main area is under sugarcane 40%, rice 17% and 6.49% Follow land (From field observation it was found that this area includes the harvested rice field area and kept ready for sowing wheat). In Saharanpur district the Mango occupies 14%.
International Conference of GIS-Users, Taza GIS-Days, May 23-24, 2012 Proceeding Book
500
The classification accuracy analysis showed that the overall classification of crop inventory during kharif season 2005 was 91% and for Rabi season 2006 the accuracy assessments was 93%. Area under various land use /land cover was derived from the classified image of rabi season (fig.3). Wheat is the main cereal grown in Saharanpur district 50.89%, sugarcane 6.03% and fallow land 9.25 %. b. Yield –NDVI relationship for Saharanpur district. In this work we used a simple model (vegetation index) in order to have a better correlation between NDVI and yield. At Saharanpur district, the dominant crop it is wheat. [4], [5] "For predicting wheat yield we used the relationship between NDVI and yield which was developed in the soil and agriculture division (IIRS-Campus)".
6.
Conclusions. The results of the study lead to the following major conclusions: o Detailed discrimination and acreage estimation of major Rabi and kharif crop seasons prepared by digital supervised classification of satellite data - IRS-1C LISS III. o The results of land use /land cover analysis reveal that of the area of Saharanpur district falling under dominant kharif season sugarcane, the total cropped area (sugarcane and paddy) in kharif was 39.2%, 11.6% respectively. In Rabi season wheat and sugarcane (50.89%, 6.03%) are the dominant crops. The major cropping system is the paddy-wheat followed by sugarcane. o IRS – 1C LISS III data can be used effectively for yield estimation of the dominant Rabi crop (wheat) using wheat yield – NDVI relationship.
Acknowledgements I would like to thank all scientists & Research Fellow of Agriculture & soil Division (IIRS-India) for helping me, from where I was completed successfully my Post Graduate Diploma. References [1] Dadhwal, V.K., Sehgal, V.K., Singh, R.P., & Rajak, D.R (2003): Wheat yield modeling using satellite remote sensing with weather data: Recent Indian experience. Mausum 54(1): 253-262. [2] Nain, A.S., Dadhwal, V.K., Singh, T.P (2004): Use of CERES-Wheat model for wheat yield forecast in central IndoGangetic plains of India. J agricultural Science (camb.), 142: 59-70. [3] Patel N.R., B.Bhattacharjee A.J., Mohammed, Tanu Priya & Saha, S.K (2006): Remote sensing of regional yield assessment of wheat in Haryana, India. International Journal of Remote Sensing, 27(19):4071-4090. [4] Patel, N.R., Manjunath, M.N., Shukla, M.R., & Pande, L.M (2004): Discrimination and empirical modeling of sugarcane and wheat crops using remote sensing and ground observations. Asian Journal of Geoinformatics, 4(4): 13-24. [5] Reddy, T.Y., & Reddi, G.H.S (2003): Principles of Agronomy. pp: 48-77. Kalyani Publishers, Ludhiana, India. [6] Saha, S.K., Patel, N.R., & Ahmed, M (2004): Evaluation of IRS-P6 LISS III and AWiFS data for rabi crop inventory: a case study of part of western Uttar Pradesh. In: Remote Sensing with Resourcesat-1: Early results, NRSA, Hyderabad. [7] Sahai, B., & Dadhwal, V.K (1990): Remote sensing in agriculture. pp. 83-98. In: J.P. Verma & A. Verma (eds.) Technology Blending and Agrarian Prosperity. Malhotra Publishing House, New Delhi. [8] Singh, R.P., Sridhar, V.N., Dadhwal V.K., Singh K.P., Navalgund, R.R (2002): Comparative evaluation of Indian Remote Sensing Multispectral Sensors data for crop classification. Geocarto International, 17(2): 5-9 [9] The Rabi crops are sown in the period between October and December and harvested in April and May (wheat, barley, peas, and mustard). [10] Kharif crops they are sown in the months of June and July and are harvested in autumn months, viz., in September and October (rice, maize, sugarcane, cotton).
Colloque International des Utilisateurs de SIG, Taza GIS-Days, 23-24 Mai 2012 Recueil des Résumés