Abstract. Introduction. Study area

Transcription

1 PROGRESSION OF RICE TRANSPLANTATIONS AND ITS IMPACT ON PRODUCTION FROM MULTI- TEMPORAL AWIFS DATA Abstract The information on the progression of rice transplantations is of paramount importance for optimum utilization of water resources, disease and pest forecast, damage assessment due to natural calamities, to aid in extension activities, input management and yield predictions for planning of procurement and storage facilities. The time of transplantation has a profound influence on the final grain yield and hence, a reliable and timely generation of data base on spatial and temporal progression of rice crop transplantation assumes importance. The potential of IRS-P6 Resourcesat satellite having 5 day revisit period with imaging capability in optical and near-infra red wavelengths has been exploited to discriminate early and late transplanted paddy crop and study their phenology through the rabi season of 23-4 in West Godavari district, Andhra Pradesh. Key Words: Paddy Early and Late transplantation, Remote Sensing and NDVI. Introduction The information on the progression of rice transplantations is of paramount importance for optimum utilization of water resources, disease and pest forecast, damage assessment to natural calamities, to aid in extension activities, input management and yield predictions for making preparations for procurement and storage facilities. The time of transplantation has a profound influence on the final grain yield and hence a reliable and timely generation of database on the spatial heterogeneity/progression of rice crop transplantation assumes importance. It is well known that sustainable agricultural development can happen only if the natural resource base upon which it depends is prudently managed. The agro-eco regions are characterized by well developed crop calender and the optimum windows for appropriate management practices too are well defined for varying situations. However in the recent years, the continuous droughts, untimely rains and its manifestation on reduced water storage in reservoirs/dams is forcing farmers to take up K. V. Ramana, M.V.R.Sesha Sai, P V V Satyanarayana* and G Jayasree** National Remote Sensing Centre Balanagar, Hyderabad 5 37, India * RARS, ANGRAU, Maruteru, W G Dt. AP ** College of Agriculture, Rajendranagar, ANGRAU, Hyderabad ramana_kv@nrsc.gov.in cultivation as when the water is made available irrespective of the optimal time window. This is truer for a crop like paddy which is water intensive. The paddy crop in its traditional form of cultivation is initially raised as nursery for a month and subsequently transplanted into the main field. Due to several reasons like weather, water and labour availability the transplantations are staggered. The temporal SAR amplitude data is capable of discrimination of early and late transplanted rice crop. An attempt has been made in this study to discriminate the early and late transplanted paddy using multi temporal optical satellite data. The IRS-P6 Resourcesat-1 satellite provides data in optical, near and middle infrared wavelengths in a large range of spatial resolutions ranging from 5.8 meters to 56 (AWiFS) meters and varied swath from as small as 23 km to 7 km and having 5 days to 23 day revisit period. The advanced wide field sensor (AWiFS) having 7 km swath and 5 day revisit period has been exploited to discriminate early and late transplanted paddy and characterize their phenology using NDVI profiles. Study area The study was conducted the West Godavari district, Andhra Pradesh covering an area of 7764 km2. The area is known as one of the rice bowl districts of Andhra Pradesh. The Southern part of the district where rice is the predominant crop is nearly level to very gently sloping with moderate to poor surface drainage. Assured irrigation is available from Godavari delta canal system. The mean annual precipitation is over 9mm, which is received mostly from South- West monsoon. The district is divided into 46 mandals. The normal period of paddy transplantation starts from December 1 and extend up to 1-15th of January. The popular varieties of paddy in the district are MTU 11 and MTU 12 with the former occupying upto 65-7% of the area under rice. ISSN No: Volume 3, Issue 6, Nov

2 Data base Nine AWiFS scenes from IRS-P6 satellite were used in the current study. (Table 1) apart from Survey of India topomaps. Table 1. Database S No. Date of Pass S No. Date of Pass Dec Feb Jan Mar Jan Mar Jan Apr Feb-24 Methodology The data preparation consisted of georeferencing of all the AWiFS scenes using Landsat-ETM orthorectified data (Path/Row 142/48 and 142/49) followed by conversion of digital numbers(dn) values to physical units of in-band radiance (mw.cm2-sr) using the offset and gains given in the header of the data sets. The Gaussian maximum likelihood per-pixel classifier was used to classify the early and late transplanted paddy crop using the data sets of 3 Jan, 7 Jan, 12 Jan, 17 Jan and 19th Feb, 24. The crop that was manifested in the data of January was considered as early transplanted rice crop and that was manifested in February was classified as late transplanted rice crop. A mask of early and late transplanted paddy was prepared. Also, all the data sets were processed for NDVI extraction using NIR and red bands only for mask of early and late transplanted paddy areas. The spectral response pattern of paddy transplanted at different stages was generated. The vector layer of the mandal map was used to extract the statistics of acreage and NDVI means. paddy crop. The map showing spatial distribution of early and late transplanted paddy is shown in Figure 1f. The figure 2 depicts the progression of paddy transplantations. It could be observed that there was sharp increase in the paddy area during the period 17th Jan to 29th Feb 24. The Figure 3 depicts the total area and cumulative percentage of area under paddy crop in mandals with different proportion of paddy crop. It can be inferred that mandals having 75 1 % and 5-75% of their area under paddy crop contributes 39% and 38% respectively to the total area in the district. The areas captured by the sensor on 17th January, 24 have been considered as early transplanted crop. Areas classified as paddy subsequent to this date have been considered as late transplanted paddy. Ignoring the mandals where the total paddy area is less than 5 ha the acreage under paddy of both the categories was found to be 2.2 lakh ha covering 34 mandals of the total 46 in the district (Fig 1f and Table 2). The early transplanted paddy constitutes an area of 1.9 lakh ha (54 %) and is predominant (> 5%of the area under total paddy crop) in 16 mandals, whereas the late transplanted paddy occupies an area of.93 lakh ha ( 46 %) spread over 18 mandals. Analysis of multidate simulated WiFS from IRS- LISS-I data over parts of West Bengal revealed that it was possible to derive much useful cropping system information like cropping pattern, crop rotation etc. (5). Results and Discussion Discrimination of early and late transplanted paddy areas: The normal transplantation starts around 1th December and the crop takes 3-4 weeks period for it to manifest in the image. A close look at the Southern part of the temporal images in Fig 1 a,b,c,d and e for the study area representing the 12thDecember (fallow), 17thJanuary (early transplanted showing spectral emergence), 19th February and 24th March (Total area under paddy crop could be seen) and 12th April (Harvested fields of early transplanted paddy) respectively depicts the pre-transplantation stage to senescence of the ISSN No: Volume 3, Issue 6, Nov

3 Area ( ha) Cumulative percentage Area ( ha) Figure 1. Temporal images of West Godavari district (a,b,c,d and e) and the Spatial distribution of early and late transplanted paddy (f) 2 Late transplanted Polavaram-2, Penumantra, Attili, Pentapadu,Poduru, Tadepalligudem, Bhimadole-2, Iragavaram, Tanuku, Achanta, Yelamanchili, Lingapalem 1, Penugonda, Peravali, Nallajerla, Denduluru, Dwaraka Tirumala, ans Lingapalem-2 Chance of escaping low temp during the early stages Reduced effect of grain shattering Pest attack depending upon the breeding cycle Farmers are realizing more grain yield to the tune of.4.5 T/ha S No Dec-3 Table 2. Likely impact of early and late transplantation Mandal Names Fig 2. Progression of paddy transplantations 3-Jan-4 7-Jan-4 17-Jan-4 Early transplanted Undi, Kalla, Bhimadole, Moghalthur, Nidamarru, Bhimavaram, Palakollu, Palakoderu, Bhimadole, Ganavaparam, Veeravasaram, Undrajavaram, Nidadavole, Palacole, Chagallu, Unguturu and Polavaram 19-Feb-4 Date of observation 29-Feb-4 Progression of paddy transplantation (ha) Fig 3. Proportion of area under rice in different mandal Mandals with differnt percent of area under paddy crop (%) Area ( ha) in various strata Cumulative percentage 1-Mar-4 1. Probable affects Mar-4 Likely to face low temp in early stages resulting in reduced growth rate; Encounter Zinc def; Grain shattering resulting in yield reduction Pest attack depending upon the breeding cycle Temporal NDVI profile of early and late transplanted paddy: The availability of the time series data within the crop season at spatial resolution of 56 m associated with the swath of about 74 km is the prime advantage of wide swath of AWiFS sensor. In order to study the temporal behaviour of NDVI, three typical mandals each representing predominantly early and late transplanted paddy were selected and the mandal mean NDVI for early and late transplanted paddy masks were shown in figs. 4a and b, respectively. It can be inferred from the fig 4.a that the spectral emergence for early transplanted paddy started showing only after Jan 7th onwards whereas the late ones from 17th January onwards. The growth rate and senescence are quite different in both the stages of paddy. The rate of senescence is rapid in the late paddy than the early transplanted paddy. The mean temporal NDVI profile of the three mandals of the early as well as late is depicted in figure 4c. It is clearly evident from the figure that the growth rate in the late paddy is less that the early one and peak vegetative growth achieved in the early paddy expressed in terms of NDVI is also high. Because of the higher temperatures like to be faced by the late paddy, the senescence is faster. The overall picture is that the early transplanted is having longer duration for synthesizing more grain should reflect in higher grain yield. However during the year of this study, the early transplanted paddy faced the problem of grain shedding due to unusually low night time temperatures, reducing the yield by 1-15%. The early sown paddy occupying an area of 1.9 lakh ha in 16 mandals of the study area are prone to such natural vagaries suggesting that the delayed plantation may save the farmers of the grain yield. The likely impact of early and late transplanted paddy is given in Table 3. Production oriented survey (1) for rice crop also highlighted the problems likely to be encountered in early and late transplanted paddy crop. Several studies carried out earlier also used remote sensing ISSN No: Volume 3, Issue 6, Nov

4 NDVI NDVI data and derived index like NDVI for assessing environmental impacts on crop growth conditions (4), for estimating LAI, green biomass and yield (2; 3;5; 6; 7). Fig 4a. Temporal NDVI profile of early transplanted paddy for selected mandals Jan- Dec Jan Jan- 17-Jan- 4 4 Date 19- Feb Feb-4 1- Mar Apr- Mar-4 4 Bhimadole-E Palakoderu-E Undi-E Mean Fig 4b. Temporal NDVI profile of late transplanted paddy for selected mandals Jan- Dec Jan Jan Jan Feb-4 Date 29- Feb-4 1- Mar Apr- Mar-4 4 Attali-L Pentapadu-L Poduru-L Mean Table 3. Spatial extent of early and late transplanted paddy S No. Mandal Name Geographical Area (ha) Total Paddy Area (ha) % of total paddy Area under Area under Early paddy Late Paddy % of early paddy to total paddy % of late paddy to total paddy Mandals with Early transplantation 1 UNDI KALLA BHIMADOLE MOGALTHUR NIDAMARRU BHEEMAVARAM PALAKODERU BHIMADOLE-W GANAPAVARAM VEERAVASARAM UNDRAJAVARAM NIDADAVOLE PALACOLE CHAGALLU UNGUTURU POLAVARAM-I Mandals with Late transplantation 17 POLAVARAM PENUMANTRA ATTILI PENTAPADU ISSN No: Volume 3, Issue 6, Nov

5 21 PODURU TADEPALLIGUDEM BHIMADOLE-E IRAGAVARAM TANUKU ACHANTA YELAMANCHILI LINGAPALEM PENUGONDA PERAVALI NALLAJERLA DENDULURU DWARAKA TIRUMALA LINGAPALEM-P Conclusions The study vividly demonstrates the potential of multitemporal AWiFS data to discriminate the early and late transplanted paddy. The spatial extent under early and late transplantation was found to be 54 and 46 % respectively. The temporal NDVI profiles of early and late transplanted paddy exhibited unique patterns at each phenological stage in terms of growth and senescence. In order to make effective utilization of this information, there is a need to link to relate this information with yield and integrated with the district agricultural plans. Acknowledgments The authors would like to place on record their sincere thanks to Director, National Remote Sensing Centre, Hyderabad, India, for providing necessary facilities and encouragement during the course of investigation. vegetation condition indicesinternational Journal of Remote Sensing, 23 (6) pp [5.] Navaslgund.R.R, Parihar.J.S., Venkataratnam.l., Krishna Rao.M.V., Panigrahy.S., Chakraborty.M.C., Hebbar.K.R., Oza.M.P., Sharma.S.A., Bhagia.N and Dadhwal.V.K Early results from crop studies using IRS-1C data. Current Science. 7 (7) p. [6.] Parihar,J.S and Dadhwal,V.K., 22. Crop production forecasting using Remote sensing data: Indian experience. Paper presented at ISPRS commission VII. Symposium Resource and Environmental monitoring, Hyderabad. (3-6 Dec, 22) [7.] Rasmussen, C.E., Williams, C.K.I., Gaussian Processes for Machine Learning. The MIT Press, New York. References [1.] Annonymous, Production Oriented Survey. Published by Directorate of Rice Research, Hyderabad [2.] Gat, N., Erives,H., Fitzgerald,G.J., Kaffka,S.R., and Mass,S.J., 2., Estimating sugar beet yield using AVRIS-derived indices. Source:web: eb.pdf [3.] Groten S.M.E., NDVI crop monitoring and early yield assessment of Burkina Faso. Intl. J. Remote Sensing. 14(8), [4.] Liu, W.T., and Kogan.F. 22. Monitoring Brazilian soybean production using NOAA/AVHRR based ISSN No: Volume 3, Issue 6, Nov