|
|
|
Poster Session 2
|
Identification and Mapping of Irrigated Vegetation
using NDVI-Climatological Modeling
4. Methodology
Pathfinder A VHRR Land (PAL) 8 km data from NOAA 11 satellite was obtained from the Goddard Distributed Active Archive Center (DAAC), USA. .The data for the year 1989 was selected for the study as it was a relatively normal year with respect to Indian monsoon which greatly influences vegetation growth and dynamics in the region. The NDVI is calculated from atmospherically corrected surface reflectance from the visible (0.58 -0.68) and near infra-red (0.725 -1.100) AVHRR channels. The 10 day composite ; NDVI data has been prepared by selecting the maximum NDVI value from ..the daily data during the 10 day period. This removes the effect of clouds to give a cloud free data set. A monthly average NDVI data was prepared by calculating the: mean value for each pixel from the 10 day composite NDVI data set. As the area has different well defined growing seasons, as discussed later, seasonal average and total NDVI data sets were also prepared. These seasons include July to October (hot/humid monsoon season), November to March (dry winter season) and April to June (Summer season). Annual average and total NDVI of the area was also calculated.
The daily data on temperature and rainfall was obtained from Global Daily Summary for the years 1977-1991 produced by National climatic Data Center, USA. Mean monthly temperature and total monthly rainfall were calculated from the daily data from about 177 meteorological stations spread allover India. The average data values for the 15 years period we re then calculated. The point-location data were subsequently interpolated to prepare average mean monthly temperature and average total monthly rainfall maps of the region. These data are analyzed vis-a-vis NDVI in different growing seasons.
5. Results and Discussions
The natural vegetation in the Indian peninsula has tropical dry deciduous and moist deciduous forests mainly in central and southern parts which shed their leaves in the autumn season. Therefore, the NDV~ in these areas starts decreasing in November and again starts increasing in March. However, India also has some areas under tropical evergreen forests in the
northern Himalayas, north-eastern states and along western sea coast which show higher NDVI most of the year .
The northern Indo-Gangatic plains are the largest agricultural region in the area which has substantial irrigated agriculture. Besides this there are other agricultural areas scattered allover India having both dry land and irrigated agriculture. There are distinctively two agricultural cropping seasons locally called as Kharif and Rabi season s. Kharif season begin sin July and lasts in October. It is a hot and humid season characterized by heavy monsoonal rainfalls. Therefore, the NDVI in the area starts increasing with the sowing of Kharif crops in July and reaches its peak in the month of October (Fig. I). Then it suddenly drops in November because of ripening and harvest of the agricultural crops. Rabi season extends from November to April which is dry and relatively cold particularly in the northern India. Therefore, NDVI again starts increasing in December indicating canopy development of the Rabi season crops. It reaches its peak in February / March (Fig. 2) and then suddenly drops in April because of yellowing and harvesting of the crops. The
season from April to June is relatively hot and dry and the agricultural fields : remain almost vacant showing very low NDVI.
Fig. 1. Average NDVI image of India dl.U"ing October} 1989.
Fig. 2. Average NDVI of India dilling February 1989.
The average NDVI data for these three major seasons and annual average NDVI is being analyzed with the average monthly mean temperature and : average monthly total rainfall. In our current work we are continuing to develop, test and refine the NDVI- climatilogical method which shows a promise for automated identification and mapig and mapping of irrigated vegetation
promise for automated identification and mapping of irrigated vegetation.
References
-
Asrar, G., Fuchs, M., Kanemasu, ET. and Hatfield, J.L. 1984. &timating ~ absorbed photosynthetic radiation and leaf area index from spectral r reflectance in wheat. Agron. J.,76: 300-306.
- Dadhwal, V.K, Shah, A.K and Vora, A.B. 1994. Carbon flow through Indian
r agroecosystem: A priliminary account. In Global O1ange Studies, Scientific ,Report, ISRO-OBP-SR-42-94. ;
- Emori, Y., Yasuda, Y., Yamamoto, H. and Isaka, I. data.base of spectral ~ignature in Japan. In: Proc. Int. Symp. on Remote SensIng for (J)servatIon and Inventory of Earth Resources and the Fndangered Fnvironment, Freiburg (FRO), 2- 8 July, 1978, 22(7/ 1 ):543 -566.
- Goward, S.N., Tucker, CJ. and Dye, DO. 1985. North American vegetation
patterns observed with the NOAA-7 Advaned Very High Resolution
Radiometer. Vegetatio, 64: 3 -14.
- Kumar, M. and Monteith, J.L. 1981. Remote sensing of crop growth. In: H.
Smith (ed.), Plants and the daylight spectrum. Academic press, London, pp.
133-144.
- Price, S.D 1991. A model of regional primary production for use with course
resolution satellite data. Int. J. Remote Sensing, 12: 13 13 -133 0.
- Ruimy, A., Saugeir, B. and ~dieu, a. 1994. Methodology for estimation of
terrestrial net primary production from remotely sensed data. J. aeophy. R.,
99: 5263-5283.
- Stockle, C.O. and Kiniry, J.R. 1990. Variability in crop radiation-use efficiency
associated with vapor-pressure deficit. Field Crops Research, 25: 171-181.
|
|
|
|
|
|
|
