International Journal of Applied Earth Observation and Geoinformation

Transcription

1 International Journal of Applied Earth Observation and Geoinformation 12 (2010) Contents lists available at ScienceDirect International Journal of Applied Earth Observation and Geoinformation journal homepage: Non-point source pollution in Indian agriculture: Estimation of nitrogen losses from rice crop using remote sensing and GIS Abha Chhabra, K.R. Manjunath, Sushma Panigrahy Agriculture Forestry and Environment Group, Remote Sensing Applications Area, Space Applications Centre, Indian Space Research Organization, Ahmedabad , Gujarat, India article info abstract Article history: Received 6 October 2009 Accepted 12 February 2010 Keywords: Nitrogenous fertilizers Rice crop Leaching loss Ammonia volatilization loss Indo-Gangetic plain region The paper presents a detailed understanding of nitrogenous fertilizer use in Indian agriculture and estimation of seasonal nitrogen loosses from rice crop in Indo-Gangetic plain region, the food bowl of the Indian sub-continent. An integrated methodology was developed for quantification of different forms of nitrogen losses from rice crop using remote sensing derived inputs, field data of fertilizer application, collateral data of soil and rainfall and nitrogen loss coefficients derived from published nitrogen dynamics studies. The spatial patterns of nitrogen losses in autumn or kharif and spring or rabi season rice at 1 1 km grid were generated using image processing and GIS. The nitrogen losses through leaching in form of urea-n, ammonium-n (NH 4 -N) and nitrate-n (NO 3 -N) are dominant over ammonia volatilization loss. The study results indicate that nitrogen loss through leaching in kharif and rabi rice is of the order of 34.9% and 39.8% of the applied nitrogenous fertilizer in the Indo-Gangetic plain region. This study provides a significant insight to the role of nitrogenous fertilizer as a major non-point source pollutant from agriculture Elsevier B.V. All rights reserved. 1. Introduction Agriculture has been identified as the largest contributor of non-point source (NPS) pollution of surface and ground water systems globally (Thorburn et al., 2003). Fertilizers, which are used as important inputs in agriculture to supply essential nutrients like nitrogen (N), phosphorus (P), and potassium (K) also serve as a major non-point source pollutant. They undergo transformation through various physical, chemical, and biological processes in soil. From the beginning of green revolution, nitrogen fertilizer use has been a success story for Indian crop production. The increase in nitrogenous fertilizer application in the last four decades (0.05 Mt in to 11.7 Mt in ) has resulted in unprecedented increase in agricultural production in the northwestern India leading to food security of the country (NAAS, 2005). Recently, concerns have been raised regarding consequences of fertilizer use more particularly nitrogenous fertilizers, since fertilizer recovery efficiency of nitrogen seldom exceeds 50% and a major portion of applied fertilizer is lost from plant soil system by various soil processes. Therefore, the losses of reactive nitrogen from agricultural systems are a serious cause of concern for both economic and environmental reasons (Raghuram et al., 2007). Assessing the environmental impacts of non-point source pollutants at a localized Corresponding author. Tel.: ; fax: address: abha@sac.isro.gov.in (A. Chhabra). and regional scale is of prime importance to achieve sustainable agriculture. Rice is the most dominant food crop of India. Nitrogen is the single most important nutrient element that has a profound effect on the growth and yield of rice. Most of the Indian soils are deficient in available N, and rice crop invariably respond to the application of fertilizer-n in almost all the Indian soils (Panda et al., 2007). Among the different forms of fertilizer nitrogen, nitrate (NO 3 )is highly mobile and most susceptible to leaching, ammonium (NH 4 + ) the least and urea-n is moderately susceptible. Urea is the most common form of nitrogenous fertilizer applied to rice crop in India. When urea is applied to aerobic or flooded rice soil, it hydrolyses to ammonium carbonate by the enzyme urease, principally produced by soil microbes. Depending upon the alkalinity and ph buffering capacity of soil, ammonium carbonate decomposes and ammonia gas escapes to the atmosphere resulting in ammonia volatilization loss. Urea hydrolysis is maximum in soils with high organic carbon or added organic manure or green manure, ph around 8.0, moisture status at field capacity, and temperature around 35 C (Panda et al., 2007). The ammonium ion resulting from urea hydrolysis also undergoes biochemical oxidation (nitrification) in aerobic soil as well as at three sites in submerged rice soil system viz. top oxidized soil layer, rice rhizosphere, and flood water. The diffusion of nitrate into the anaerobic zone in the submerged soil also results to denitrification loss (N 2 O, N 2 ) of nitrate (NO 3 ) formed. Nitrate leaching is dominant when both soil nitrate content and water movement are high. The NO 3 anion is very soluble in water, /$ see front matter 2010 Elsevier B.V. All rights reserved. doi: /j.jag

2 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) not influenced by soil colloids and unless intercepted and taken up by plant roots, leach down in the soil along with irrigation or rainwater or carried away by run-off. Nitrate leaching from field soil must be carefully controlled with its associated environmental impacts. High nitrate levels in surface run-off and water percolating through the soil can pollute drinking water resources and stimulate unwanted plant and algae growth in lakes and reservoirs (Panda et al., 2007). A detailed study was made on nitrogen fertilizer use and various forms of gaseous N emissions including N 2 O, NO x and NH 4 volatilization from rice based cropping systems in India (Adhya et al., 2007). Ammonia volatilization is not only a major loss of N primarily from rice cultivation but also a cause of environmental pollution. From the atmosphere NH 3 is washed out by clouds and redeposited on the terrestrial ecosystem. In the atmosphere, it is oxidized to N 2 O which is a potential greenhouse gas and responsible for the destruction of stratospheric ozone layer Objectives of the study Various studies have been carried out to understand nitrogen dynamics in soil with varying management practices for applied nitrogenous fertilizer. However, soil type is one of the most important factors in affecting fertilizer use efficiency and soil productivity. The spatial patterns of extent and intensity of different forms of nitrogen losses in soil and their variations with soil characteristics need to be captured using remote sensing and GIS tools. Indo-Gangetic plain (IGP) is the agriculturally dominant region and also known as the major food bowl of the Indian sub-continent. Rice is the dominant food crop of Indo-Gangetic plain region occupying about 32% of the gross cropped area and consumes 49% of the total agricultural nitrogenous fertilizer used in the region. Urea is the major source of fertilizer-n for rice crop in Indian agriculture. The present study was aimed at a spatio-temporal analysis of agricultural nitrogenous fertilizer use in India and a detailed estimation of various forms of nitrogen losses from applied urea N-fertilizer in rice crop in the Indo-Gangetic plain states using remote sensing and GIS tools, other collateral and field data. The major objectives of the study include: i) Statistical analysis of N-fertilizer consumption in India and deriving the spatio-temporal patterns at district level by linking the administrative boundaries in GIS. ii) Estimation of extent and intensity of different forms of nitrogen losses (leaching and volatilization) from applied N-fertilizer in kharif and rabi rice crops for different soil types in the Indo- Gangetic plain (IGP) states using remote sensing derived inputs and GIS. 2. Materials and methods 2.1. Study area The detailed spatial analysis of nitrogenous fertilizer use in Indian agriculture was carried out at the district level. However, a detailed estimation of nitrogen losses through leaching and volatilization from applied N-fertilizer in kharif and rabi rice crops was carried out only for Indo-Gangetic plain region of India which includes five states viz. Punjab, Haryana, Uttar Pradesh, Bihar and West Bengal (Fig. 1) Data used Statistical data A detailed district level statistical database of nitrogenous fertilizer use was prepared using fertilizer statistics (FAI, 2004) available online at A detailed database of location, soil texture, ph, organic carbon, season, crop, year, applied N fertilizer (kg ha 1 ), N-fertilizer type, N uptake (kg ha 1 ), N use efficiency (percent), N uptake efficiency (percent), apparent N recovery (percent), leaching losses (urea-n, NH 4 -N, NO 3 -N percent) NH 4 volatilization and gaseous N 2 O emission was prepared using published literature studies (Appendix A). The study database included published nitrogen dynamics studies of different crops viz. rice, wheat, maize and cotton. The present study was focused on rice crop only as few studies on N-losses with different soil types were available, as compared to very limited data available for other crops. Fig. 1. Indo-Gangetic plain region of India constituting five study states.

3 192 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) Remote sensing, GIS and other spatial data Remote sensing derived seasonal rice maps of the study states (Fig. 2a) (Manjunath et al., 2006) were used as an input for seasonal rice crop area at 1:1 km in each of the study states. The Indian soils have been classified into eight texture classes (SOI, 1978), of which Indo-Gangetic plain states are characterized by five texture classes viz. sand, sandy loam, loam, clay loam and clay (Fig. 2b). The seasonwise rice cropped area under different texture classes in each of the study state was estimated from soil texture map and remote sensing derived seasonal rice maps using image processing in PCI Geomatica version 9.0. The field database of per hectare nitrogenous fertilizer applied in kharif and rabi seasons rice crop with their soil texture information at different locations in each of the study state was used. This database forms a part of a methane emissions inventory studies for rice ecosystems in India (Manjunath et al., 2009). The N-fertilizer applied field data in form of urea or DAP was converted in form of N-applied and then used as an input for this study. ARC/INFO GIS coverage of administrative boundaries of all districts was used to link statistical database and deriving spatial patterns of nitrogenous fertilizer use. Maps representing estimated nitrogen losses from leaching and volatilization from kharif and rabi rice crops were generated through image processing. GIS coverage of administrative boundaries of Indo-Gangetic plain states was laid over the study region. The mean annual rainfall pattern (Fig. 2c) derived from climatic normal was overlaid to infer the impacts of fate of nitrogen losses to soil or ground water. The rainfall pattern was studied by classifying into five zones, viz. <60, , , , and >200 cm. The estimated nitrogen losses were studied with reference to the rainfall zones to infer the likely effect of rainfall and pollution potential arising from nitrogenous fertilizer use Methodological approach The methodology involved in this study included the use of district level statistical data of nitrogenous fertilizer use and its linkage in GIS to derive the spatial patterns. The nitrogen losses in form of leaching and ammonia volatilization were estimated using remote sensing derived seasonal rice crop area, collateral data of soil map and field data of fertilizer use intensity. The spatial patterns of nitrogen losses were derived using image processing software PCI Geomatica version 9.0. The outputs were generated as maps at 1-km grid size for the entire study region including all five Indo-Gangetic plain states. The methodological approach may be classified into statistical and remote sensing based approach as discussed in detail in the following sub-sections Statistical and GIS linkage A statistical approach was used to study the nitrogenous fertilizer use pattern among different districts of India. The spatial framework of GIS coverage of administrative boundaries of all districts was created using ARC/INFO ver 8.3 with Albers Conical Equal Area projection for further analysis and deriving spatio-temporal patterns. Rice is grown as a major autumn or kharif crop in all the five Indo-Gangetic plain study states. However, spring or rabi rice crop is grown predominantly in West Bengal and few parts of Bihar. Puddling is a common agricultural management practice of rice cultivation in the study area. The coefficients of nitrogen losses in form of leaching and ammonia volatilization from kharif and rabi rice crops were derived from database of published field studies of nitrogen dynamics and logical decision rules. These coefficients were derived as percentage of nitrogen losses for different leaching forms (urea-n, NH 4 -N, and NO 3 -N) and ammonia volatilization in different soil types for kharif and rabi rice crop. As limited observations were available for nitrogen losses in different soil tex- Fig. 2. (a) Rice map of Indo-Gangetic plain states. (b) Soil texture map of Indo- Gangetic plain states. (c) Mean annual rainfall pattern of Indo-Gangetic plain states.

4 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) Table 1 Coefficients used for estimating different N losses from rice crop in Indo-Gangetic plain states. Soil texture Kharif rice Rabi rice leaching N loss (%) volatilization loss (%) leaching N loss (%) volatilization loss (%) Urea-N NH 4-N NO 3-N Urea-N NH 4-N NO 3-N Sand a a a a Sandy loam Clay loam not present. a Rabi rice crop not grown. tures, leaching and ammonia volatilization coefficients computed for sandy loam and clay loam texture classes were also applied for loamy and clay soils, respectively. A summary of coefficients used for this study is presented in Table Remote sensing approach The remote sensing derived seasonal rice maps from multisource satellite data were applied in conjugation with soil map to compute season-wise rice area for sand, sandy loam, loam, clay loam and clay soils in Indo-Gangetic plain study states. The mask for each of the study state was prepared and seasonal crop area in each texture class was derived under each state mask in image processing PCI Geomatica ver 9.1. The few regions in states of West Bengal and Bihar where rice is grown in both kharif and rabi seasons were considered under rabi crop area only for the purpose of estimation of seasonal nitrogen losses. The magnitude or total quantum of nitrogen losses (expressed in 000 tonnes) for rice crop in different soil classes was estimated using remote sensing derived seasonal rice cropped area, season-wise field data of N-fertilizer applied to rice for each soil type and derived coefficients of nitrogen losses from published studies. The estimated quantum nitrogen losses were further aggregated at the state level. The intensity of different forms of nitrogen losses (leaching and volatilization) from rice crop were estimated using derived coefficients of nitrogen losses in different soil types from published studies and field data of N fertilizer applied (kg ha 1 ) in kharif and rabi seasons. The weighted average nitrogen loss in kharif and rabi rice grown in different soil types in each study state were estimated to arrive at state level intensity of nitrogen loss by leaching and ammonia volatilization. The annual rainfall pattern map derived from climatic normal was overlaid to infer the impacts of fate of nitrogen losses to soil or ground water. The rainfall pattern was classified into five zones, viz. <60, , , , and >200 cm. The estimated nitrogen losses were analyzed with reference to the rainfall zones to infer the likely effect of rainfall and pollution potential arising from nitrogenous fertilizer use. The spatial patterns of leaching and volatilization losses in different soil textures for kharif and rabi crop areas in the Indo-Gangetic plain states were prepared in image processing PCI Geomatica ver 9.1 at 1-km grid size. These spatial patterns were then exported in ArcGIS for generating maps Fig. 3. Methodology flow diagram (the shaded area depicts the methodology adopted for studying district level spatial patterns of N fertilizer use while the remaining part indicates the approach for N-loss estimation for IGP states).

5 194 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) Fig. 5. Spatial pattern of nitrogen leaching loss from kharif rice in Indo-Gangetic plain states. Fig. 4. District-level nitrogen fertilizer use intensity ( ). using state level GIS coverage of Indo-Gangetic plain region. Fig. 3 presents a broad methodology flow diagram of the study. 3. Results and discussion 3.1. Statistical analysis Fertilizer statistics indicate that N-fertilizer consumption account to 64% of the all-india NPK consumption for the period to The detailed spatial analysis indicated distinct variations in N-fertilizer use with high application intensity in few districts of Andhra Pradesh, Haryana, Punjab, Uttar Pradesh, and Tamil Nadu (Fig. 4). Panipat and Karnal districts (Haryana), Udham Singh Nagar (Uttarakhand), Varanasi (Uttar Pradesh), and Nizamabad (Andhra Pradesh) have high N fertilizer use intensity above 200 kg ha 1. Ludhiana and Ferozpur districts (Punjab), West Godavari and Krishna districts (Andhra Pradesh), Howrah (West Bengal), and Tirunelveli (Tamil Nadu) are among the districts with N-fertilizer use intensity in the range of kg ha 1. The majority of districts across various states have application intensity in the range of 1 50 kg ha 1. sity of N-fertilizer application varies from 75 kg ha 1 in clay soil (West Bengal) to 166 kg ha 1 in clay loam (Punjab). The estimated leaching loss intensity among the states varies with the N-fertilizer application and coefficients of N-loss for different soil types. The nitrogen lost through leaching in form of nitrate nitrogen (NO 3 - N) is the most dominant in loam and sandy soils of Punjab and Haryana ( 29% of applied N-fertilizer). Coarse textured soil such as sand, and sandy loam encourages higher leaching than fine textured soil like clay and clay loam. Leaching of nitrates is more likely to occur in sandy soils but it takes place in fine textured soils also (Rao and Puttanna, 2000). Higher the percolation of water, higher is the leaching loss of nitrogen. In sand, loam and sandy loam soils, urea moves to deeper soil layers as compared to clay loam soil. Due to rapid transformation of urea-n to NH 4 + and NO 3 -N 3.2. Nitrogen losses from rice crop in Indo-Gangetic plain states Rice is a dominant crop in Indo-Gangetic plain region with total crop area of 22.2 Mha including kharif and rabi rice crops. Fertilizer- N is one of the key inputs in rice production and accounts for about 67% of the total amount of fertilizers applied to rice crop. The nitrogen losses in form of leaching and ammonia volatilization in kharif and rabi crop were quantified and spatial patterns of estimated nitrogen losses in the study states were generated at 1-km grid Leaching nitrogen loss The different forms of leaching losses were estimated for kharif rice crop in all five Indo-Gangetic plain study states. The inten- Fig. 6. Spatial pattern of nitrogen leaching loss from rabi rice in Indo-Gangetic plain states of Bihar and West Bengal.

6 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) Table 2 Estimated quantum of leaching nitrogen losses from kharif rice in Indo-Gangetic plain states. State Soil texture Kharif crop area (000 ha) N fertilizer applied (000 tonnes) Leaching loss (000 tonnes) Urea-N NH 4-N NO 3-N Total Punjab Sand Loam Clay loam Total Haryana Sand Loam Clay loam Sandy loam Total Uttar Pradesh Sandy loam Loam Clay loam Clay Total Bihar Sandy loam Loam Clay loam Total West Bengal Sandy loam Loam Clay Total forms, no urea-n is detected in the soil profile in sand or sandy loam soils. Rice is mostly grown in lowland conditions in the Indo- Gangetic plain region. In fine textured clay and clay loam flooded rice soils, nitrification proceeds at a slow rate and with reduced percolation the possibility of leaching of NO 3 -N beyond root zone is small, therefore leaching loss is confined to urea-n form only (Singh et al., 2006). The estimated intensity of leaching N-loss varies with soil type; from 9.2 kg ha 1 or 12.3% in clay soil in West Bengal to 64.1 kg ha 1 or 42.7% of the applied fertilizer in loam soil in Punjab. The weighted average nitrogen loss (including urea- N, NH 4 -N, and NO 3 -N) from kharif rice by leaching varies from 28.9 kg ha 1 in Bihar to 51.4 kg ha 1 in Punjab. The spatial pattern of leaching N-loss in kharif rice is presented in Fig. 5. The intensity of leaching N-loss in rabi rice in Bihar is higher as compared to kharif rice leaching losses due to higher N-fertilizer application. The weighted average nitrogen loss intensity from rabi rice in Bihar and West Bengal is estimated at 39.7 and 39.8 kg ha 1, respectively. The spatial pattern of leaching N-loss in rabi rice is presented in Fig. 6. The magnitude of N-loss through leaching from kharif rice crop in the study states varies with the crop area and N-fertilizer application intensity in each soil type. The estimated amount range from 0.05 Mt or 31.5% of the applied N-fertilizer in Haryana (1.2 Mha crop area) to 0.25 Mt or 37% of the applied N-fertilizer in Uttar Pradesh (6.1 Mha crop area). The detailed estimates of quantum of leaching loss from kharif rice in study states are given in Table 2. The parts of Punjab and Haryana with sandy soil and receiving less than 60 cm annual rainfall are most likely to contribute leached nitrogen to ground water due to high percolation rates in coarse textured soil. This is in contrast to lesser percolation of lost nitrogen in deeper layers of fine textured clay loam soil in parts of Punjab, Haryana and Uttar Pradesh receiving less than 100 cm rainfall. However, the large nitrogen loses through leaching in dominant sandy loam soil with high rainfall (more than 150 cm) are more likely to contribute through surface run-off and transport by water and getting deposited in nearby water bodies. Since the recovery of fertilizer-n is generally below 50%, the nitrate contamination of ground water can take place if water percolating from a large area converges to a point beneath the soil due to hydrological or other reasons (Rao and Puttanna, 2000) Ammonia volatilization loss The nitrogen loss in form of ammonia volatilization is less as compared to leaching loss. Ammonia volatilization from applied N fertilizers and manure may change across various agro-ecosystems, as ammonia lost to the atmosphere is affected by the rate of N-fertilizer application, modified forms of urea, source of N and moisture content of the soil (Sharma et al., 1992). Besides, the magnitude of ammonia loss also depends on wind speed, temperature, rainfall, ammonical-n concentration, ph and cation exchange capacity of soil (Freney et al., 1983). Temperature has a complex role in ammonia volatilization loss but it is difficult to model its role over a large area. The present study reports large area estimation of nitrogen losses and the role of temperature in ammonia volatilization have been considered constant in the model. The results of this study indicated that ammonia volatilization loss in kharif season varies from 4.9 kg ha 1 in sandy loam soil in Bihar to 11.6 kg ha 1 in clay loam soil in Punjab. The soil texture indirectly plays a significant role in ammonia loss. In coarse textured soils, NH 4 -N is readily transported to sub-surface layers along with percolating water. As NH 4 -N placed at depth cannot move back to soil surface due to downward flux of percolating water under repeatedly irrigated rice culture, highly permeable soils like sand or sandy loam under wetland rice conditions do not favour substantial losses of N via ammonia volatilization (Singh et al., 2007b). Ammonia loss is higher in fine textured clay or clay loam soil due to availability of moisture content in soil. Among the states, the estimated weighted average nitrogen loss from kharif rice by ammonia volatilization range from 5.4 kg ha 1 in Bihar to 9.4 kg ha 1 in Haryana. The spatial pattern of kharif rice ammonia volatilization losses in Indo- Gangetic plain states are shown in Fig. 7. The intensity of ammonia volatilization loss from rabi rice in Bihar and West Bengal is estimated at 10.3 and 10.9 kg ha 1, respectively. The spatial pattern of ammonia volatilization losses from rabi rice is shown in Fig. 8. The magnitude of estimated ammonia volatilization loss varies with rice area and soil type. The amount of ammonia volatilization N-

7 196 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) Fig. 7. Spatial pattern of ammonia volatilization loss from kharif rice in Indo- Gangetic plain states. loss in kharif rice varies from 0.2 thousand tonnes in clay soil to 25.1 in loam soil in Uttar Pradesh. Among the states, the ammonia volatilization N-loss from kharif rice varies from 0.04 Mt from rice area of 6.1 Mha in Uttar Pradesh, followed by 0.03 Mt from 3.5 Mha rice area in Punjab. Table 3 presents detailed estimates of ammonia volatilization loss by soil type in kharif rice for each of the Indo-Gangetic plain states Total nitrogen loss Leaching N-losses are dominant over ammonia volatilization losses in the Indo-Gangetic plain study states. The weighted average nitrogen loss by leaching from kharif and rabi rice crops in the study states range from 31.3 kg ha 1 in Bihar to 51.4 kg ha 1 in Punjab. However, the weighted average nitrogen loss by ammonia volatilization ranges from 6.5 kg ha 1 in Bihar to 9.4 kg ha 1 in Punjab. A comparative account of estimated nitrogen losses in the study states is shown in Fig. 9. The results of this study also indicate high intensity of nitrogen loss through leaching in kharif rice Fig. 8. Spatial pattern of ammonia volatilization loss from rabi rice in Indo-Gangetic plain states of Bihar and West Bengal. for coarse textured sand and sandy loam soils of Haryana and Punjab. The leached nitrogen from these regions with high N-fertilizer application intensity and assured irrigation or rainfall most likely contributes to ground water due to high percolation rates in coarse textured soil. There are increasing evidences of nitrate pollution in ground waters in Punjab and Haryana states of the Northern India (Rao and Puttanna, 2000). The degradation of ground water quality by N-fertilizer use in agriculture is essentially a nitrate leaching problem. The concentration of nitrate-n in the soil profile and quantity of water passing through the soil profile are the two main factors controlling the nitrate leaching in ground water (Singh et al., 2007c). A detailed understanding of the forms and pathways of nitrogen losses is therefore important to increase the nitrogen use efficiency, which can be translated into economic yields. Table 3 Estimated quantum of ammonia volatilization loss from kharif rice in Indo-Gangetic plain states. State Soil texture Kharif crop area (000 ha) N fertilizer applied (000 tonnes) NH 3 volatilization loss (000 tonnes) Punjab Sand Loam Clay loam Total Haryana Sand Loam Clay loam Sandy loam Total Uttar Pradesh Sandy loam Loam Clay loam Clay Total Bihar Sandy loam Loam Clay loam Total West Bengal Sandy loam Loam Clay Total

8 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) Fig. 9. Estimated weighted average nitrogen loss intensity from rice crop (including kharif and rabi rice) in Indo-Gangetic plain states. 4. Conclusion Fertilizer nitrogen has been the key input in augmenting India s agricultural production in past three decades leading to the food security in the country. Besides, nitrogenous fertilizer also serves as a major non-point source pollutant from agriculture since nitrogen use efficiency by crops seldom exceeds 50%. This study presents a detailed understanding of N-fertilizer use and quantification of seasonal N-losses from rice crop using remote sensing derived inputs. The detailed spatial analysis indicated distinct variations in N-fertilizer use with high application intensity in few districts of Andhra Pradesh, Haryana, Punjab, Uttar Pradesh, and Tamil Nadu. A remote sensing and GIS approach was developed to estimate nitrogen losses in form of leaching and volatilization from applied nitrogenous fertilizer to rice crop during kharif and rabi seasons in Indo-Gangetic plain region at 1-km grid size. The remote sensing derived seasonal rice maps in conjugation with coefficients from published nitrogen dynamics studies, collateral data of soil type and rainfall map, and field data of nitrogen fertilizer use intensity were used as inputs to estimate and derive spatial patterns of different forms of nitrogen losses in rice crop in the study states. The results of the study indicated that leaching nitrogen losses in kharif and rabi rice dominate ammonia volatilization loss in Indo- Gangetic plain states. Nitrogen loss through leaching in form of nitrate nitrogen (NO 3 -N) is the most dominant in sandy soils of Punjab and Haryana ( 29% of applied N-fertilizer). The weighted average nitrogen loss by leaching from kharif and rabi rice crops in the study states range from 31.3 kg ha 1 in Bihar to 51.4 kg ha 1 in Punjab. However, the weighted average nitrogen loss by ammonia volatilization ranges from 6.5 kg ha 1 in Bihar to 9.4 kg ha 1 in Punjab. The total N-loss (leaching and ammonia volatilization) from rice crop in the study states was estimated as Punjab 0.2 Mt, Haryana 0.05 Mt, Uttar Pradesh 0.30 Mt, Bihar 0.18 Mt and West Bengal 0.30 Mt. The study results indicate that nitrogen loss through leaching in kharif and rabi rice is of the order of 34.9% and 39.8% of the applied N-fertilizer in the Indo-Gangetic plain region, which is a major food growing region in the South Asia. The high leaching intensity in coarse textured sand and sandy loam soils of Haryana and Punjab may likely result in nitrate pollution of ground water with downward movement of leached nitrogen with percolating water. Further detailed studies with field measurements and use of remote sensing data in a simulation modeling environment would aid in validation and large area estimation of non-point source pollution from agricultural sources with high certainty. Acknowledgements This study was carried out as part of Environmental Impact Assessment of Agricultural System project under Earth Observation Applications Mission Programme of DOS-ISRO (EO-RSMET support to International protocol/conventions). Authors are thankful to Dr. R.R. Navalgund, Director, SAC for encouragement received to undertake this study. We sincerely acknowledge the critical suggestions and guidance provided by Dr. J.S. Parihar, Deputy Director, Remote Sensing Applications Area, SAC. The Indiastat team ( is thankfully acknowledged for providing online fertilizer and agriculture statistics. Authors wish to sincerely thank two anonymous reviewers for their critical and useful suggestions for improving this manuscript.

9 Appendix A. Summary of soil nitrogen dynamics studies for different crops in India. Location ST Soil ph Soil OC Crop Season Year App. fert (N kg/ha) NU (kg/ha) NU eff. (%) App. N recovery (%) Pathways of N loss Leaching losses (%) Urea-N NH 4 -N NO 3 -N Ammonia volat. loss N 2 O emission IARI, New Delhi SL Wheat Rabi , 120, 240 CL Kharif , 60, 120, 240 CL Rabi , 100, 150 Jain et al. (1981) CL Rabi , 150, 225 Hoshiarpur, Punjab SL 8 Maize rainfed Kharif (broadcast, drill, 2 splits) 90.3, 92.7, 90.7 Sharma and Singh (1982) LS 8 40 (broadcast, 66.4, 71.3, 70.1 drill, 2 splits) Himalayan region UP SL Maize Kharif (PU) 68.9 Pal (1996) Cuttack 40 (PU) 18 Acharya and Sharma (2008) Ag. Univ Kanpur (lab expt) SL Singh et al. CL (2000) Orissa Reg.Res Station, StL-L Rice Rabi 76 (PU) Jena et al. (2003) Ranital Orissa Reg.Res Station, Ranital StL-L Rice Rabi 114 (PU) Jena et al. (2003) TN Rice Research Institute, Thanjavur CL Rice Summer (February May) Velu and Ramanathan (1996) Rice Kuruvai (July October) Rice Thaladi (November February) Assam, Research farm Agi Univ Jorhat PAU Ludhiana SL Rice Kharif SL Rice Ahmed et al. Pantnagar Univ CL Rice rainfed Pantnagar Univ CL Rice rainfed ds rice tp rice ds rice tp rice Reference Singh et al. (2007a) Kharif Khan et al Kharif Khan et al Puddled lowland SL Rice Panda (2005) PAU Punjab SL Soil column Intermittent flooding Shibu and Ghuman (2001) study Continuous flooding Rice Wet soil Katyal (2001) Dry soil Wheat Before irrig Katyal (2001) After irrig A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010)

10 Location ST Soil ph Soil OC Crop Season Year App. fert (N kg/ha) NU (kg/ha) NU eff. (%) App. N recovery (%) Pathways of N loss Leaching losses (%) Urea-N NH 4 -N NO 3 -N Ammonia volat. loss N 2 O emission Bhuwaneshwar SCL Rice Kharif Misra et al. (1995) Fine 0.4 Rice flooded Pathak (1999) IARI New Delhi SL Purakayastha and Chhonkar Pantnagar SCL Rice Tripathi et al. (1997) PAU (lab expt) SL Rice Kharif Singh et al. CL Singh et al. SL Wheat Rabi CL Kanpur (alluvial soils OGP) SL Basdeo and Gangwar (1976) L Basdeo and Gangwar (1976) CL 2.7 Parbhani (Maharashtra) C 8 Cotton Kharif Shinde et al. (1982) Agri univ. Samastipur Bihar CL 22.6 (cu. Loss under FC) Singh and Prasad (1992) 20.1 (cu. loss under flooded condition) ST: soil texture, OC: organic carbon, App. fert.: apparent fertilizer applied (urea), PU: prilled urea, NU: nitrogen uptake, NU eff.: nitrogen use efficiency, App. N recovery: apparent N recovery, Ammonia volat. loss: volatilization, N 2O emission: Gaseous N 2O emission, SL: sandy loam, CL: clay loam, StL: silt loam, L: loam, cu. Loss: cumulative loss, FC: field condition, SCL: sandy clay loam. Reference A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010)

11 200 A. Chhabra et al. / International Journal of Applied Earth Observation and Geoinformation 12 (2010) References Acharya, C.L., Sharma, A.R., Integrated input management for improving nitrogen use efficiency and crop productivity. Indian Journal of Fertilizers 4, Adhya, T.K., Pathak, H., Chhabra, A., N-fertilizers and gaseous-n emission from rice-based cropping system. In: Abrol, Y.P., Raghuram, N., Sachdev, M.S. (Eds.), Agricultural Nitrogen Use and Its Environmental Implications. International Publishing House Pvt. Ltd., New Delhi, pp Ahmed, P., Medhi, B.D., Singh, A.K., Effect of organic and inorganic sources of nitrogen on ammonia volatilization and yield of transplanted rice. Journal of Indian Society of Soil Science 54, Basdeo, Gangwar, B.R., Studies on losses of ammonia by volatilization from nitrogenous fertilizers applied to soils. Journal of Indian Society of Soil Science 24, FAI, Fertilizer Statistics, , The Fertilizer Association of India, New Delhi, Available online at Freney, J.R., Simpson, J.R., Denmead, O.J., In: Freney, J.R., Simpson, J.R. (Eds.), Gaseous Loss of Nitrogen from Plant Soil Systems. Martinus Nijhoff/Dr W. Junk Publisher, The Hague, Netherlands, pp Jain, J.M., Sarkar, M.C., Deori, M.L., Interaction of mechanisms of gaseous losses of nitrogen in upland soils. Journal of Indian Society of Soil Science 29, Jena, D., Misra, C., Bandyopadhyay, Effect of prilled urea and urea super granules on dynamics of ammonia volatilization and nitrogen use efficiency of rice. Journal of Indian Society of Soil Science 51, Katyal, J.C., Fertilizer use situation in India. Journal of Indian Society of Soil Science 49, Khan, U., Mishra, B., Pachauri, P., Kumar, Y., Effect of integrated nitrogen management on yield and nitrogen nutrition of irrigated rice. Journal of Indian Society of Soil Science 76, Manjunath, K.R., Panigrahy, S., Kundan Kumari, Adhya, T.K., Parihar, J.S., Spatiotemporal modeling of methane flux from the rice fields in India using remote sensing and GIS. International Journal of Remote Sensing 27, Manjunath, K.R., Panigrahy S., Adhya, T.K., Beri, V., Rao, K.V., Parihar, J.S., Methane Emission inventory from Indian rice ecosystems using remote sensing, GIS and field observations, Scientific report, Space Applications Centre (ISRO), Ahmedabad, SAC/AFEG/AMD/EIAA/SN/03/2008. Misra, R.K., Jena, D., Bandopadhya, K.K., Schepers, J.S., Urea hydrolysis and ammonia volatilization from some urea based fertilizers applied to rice. Journal of Indian Society of Soil Science 43, NAAS, Policy options for efficient nitrogen use, Policy Paper no. 33, 2005 National Academy of Agricultural Sciences, New Delhi. Pal, S.S., Effect of modified urea formulations on N-use efficiency in maize. Journal of Indian Society of Soil Science 44, Panda, D., Chemistry of nitrogen transformation in submerged soil and scientific management of urea fertilizer for higher rice productivity. Journal of Indian Society of Soil Science 53, Panda, D., Kundu, D.K., Ghosh, A., Patra, D.D., Nitrogen use efficiency in rice ecosystems. In: Abrol, Y.P., Raghuram, N., Sachdev, M.S. (Eds.), Agricultural Nitrogen Use and its Environmental Implications. I.K. International Publishing House Pvt. Ltd., New Delhi, pp Pathak, H., Emissions of nitrous oxide from soil. Current Science 77, Purakayastha, T.J., Chhonkar, P.K., Evaluation of ammonia volatilization from compacted urea-micronutrient fertilizers applied in rice (Oryza sativa L.) grown on a typic haplustept. Journal of Indian Society of Soil Science 54, Raghuram, N., Sachdev, M.S., Abrol, Y.P., Towards and integrative understanding of reactive nitrogen, in: Yadav, R.L., Singh, P., R. Prasad, Ahlawat, I.P.S. (Eds.), Fifty years of Agronomic Research, New Delhi, pp Rao, E.V.S., Puttanna, K., Nitrates, agriculture and environment. Current Science 79, Sharma, B.D., Singh, R., Distribution of nitrate-n and nitrogen uptake by rainfed maize as affected by methods of application of urea. Journal of Indian Society of Soil Science 30, Sharma, S.K., Kumar, V., Singh, M., Effect of different factors on ammonia volatilization losses in soils. Journal of Indian Society of Soil Science 40, Shibu, M.E., Ghuman, B.S., Leaching of nitrogen applied as urea from undisturbed alluvial soil columns under continuous and intermittent flooding. Journal of Indian Society of Soil Science 49, Shinde, J.S., Varade, S.B., Rao, G.R., Bharambe, P.R., Evaluation of percolation and leaching loss of nitrogen applied through urea in a vertisol. Indian Journal of Agricultural Sciences 52, Singh, K.N., Prasad, B., Volatilization loss of ammonia as influenced by integrated nutrient management in calcareous soils. Journal of Indian Society of Soil Science 40, Singh, B., Upadhyay, R.M., Srivastava, O.P., Rate of leaching and downward movement of urea-n and NH 4 + -N in fine and coarse textured soils. Journal of Indian Society of Soil Science 48, Singh, V., Singh, Y., Singh, B., Meelu, O.P., Distribution and leaching losses of applied urea-n in sandy loam and clay loam soils under wetland and upland moisture regimes. Journal of Indian Society of Soil Science 54, Singh, Y., Gupta, R.K., Singh, B., Gupta, S., 2007a. Efficient management of fertilizer nitrogen in wet direct-seeded rice (Oryza sativa) in north-west India. Indian Journal of Agricultural Sciences 77, Singh, B., Singh, Y., Gupta, R.K., 2007b. Nitrogen transformations and flows in agroecosystems-an overview. In: Abrol, Y.P., Raghuram, N., Sachdev, M.S. (Eds.), Agricultural Nitrogen Use and Its Environmental Implications. I.K. International Publishing House Pvt. Ltd., New Delhi, pp Singh, B., Prakasa Rao, E.V.S., Singh, Y., Puttanna, K., 2007c. Nitrate pollution of ground water vis-à-vis nitrogen fertilizer use in India. In: Abrol, Y.P., Raghuram, N., Sachdev, M.S. (Eds.), Agricultural Nitrogen Use and Its Environmental Implications. I.K. International Publishing House Pvt. Ltd., New Delhi, pp SOI, Survey of India, Soil texture map, Thorburn, P.J., Biggs, J.S., Weier, K.L., Keating, B.A., Nitrate in ground waters of intensive agricultural areas in coastal northeastern Australia. Agriculture, Ecosystems and Environment 94, Tripathi, A.K., Singh, T.A., Singh, M., Leaching losses and use efficiency of N in rice as influenced by modified gypsum urea. Journal of Indian Society of Soil Science 45, Velu, V., Ramanathan, K.M., Nitrogen balance studies in rice wetland soil. Journal of Indian Society of Soil Science 44, Abha Chhabra, Ph.D (Environmental Sciences) is working as Scientist at Remote Sensing applications Area of the Space Applications Centre (ISRO), Ahmedabad since She specializes in remote sensing applications in the field of Environmental studies. Her PhD research work was focused on assessment and understanding the impacts of landuse/landcover changes on terrestrial carbon cycle for India. Currently, she is working on remote sensing applications for Environmental Impact Assessment studies of agriculture. She was member of the Scientific Steering committee of the IGBP-IHDP Landuse/landcover change project during Presently, she is a member of the Scientific Steering committee of the Global Land Project (GLP), a joint project of International Geosphere Biosphere Programme (IGBP) and International Human Dimensions Programme on Global Environmental Change (IHDP). K.R. Manjunath received the M.Sc. degree in Agriculture with specialization in crop physiology from University of Agricultural sciences, Bangalore, India in He is working as scientist at Space Applications Centre, Indian Space Research Organization, Ahmedabad since He specializes in remote sensing applications in the field of Agriculture, particularly crop acreage estimation, crop condition assessment, crop yield modeling, and evaluation of agricultural system sustainability. His current research contributions include Environment Impact Assessment studies of Agriculture. Mr. Manjunath is a member of many professional societies of remote sensing, GIS and meteorology. Dr Sushma Panigrahy, a Ph.D in Applied Botany from Indian Institute of Technology, Khargpur, India, joined the Space Applications Centre (ISRO), Ahmedabad as senior scientist in She specializes in remote sensing applications in the field of agriculture, particularly the rice systems in India. She has extensively worked on investigating SAR interaction with rice canopy using data from various sensors, mainly RADARSAT SAR. Currently she is working as Group Director, Agriculture, Forestry and Environment Group, Remote Sensing Applications Area at SAC, Ahmedabad.