scholarly journals RICE AREA ESTIMATION USING PARAMETERIZED CLASSIFICATION OF SENTINEL 1A SAR DATA

2019 ◽  
Vol XLII-3/W6 ◽  
pp. 141-147 ◽  
Author(s):  
M. G. Raman ◽  
R. Kaliaperumal ◽  
S. Pazhanivelan ◽  
B. Kannan
Keyword(s):  
Satellite Data ◽  
Tamil Nadu ◽  
Accuracy Assessment ◽  
Kappa Index ◽  
Transplanted Rice ◽  
Cropping Season ◽  
Rice Area ◽  
Multi Temporal ◽  
Sar Data ◽  
Crop Parameters

<p><strong>Abstract.</strong> A research study was conducted during <i>Rabi</i> 2016 (Samba season) to estimate rice area using SAR data in Tiruvarur district of Tamil Nadu. Multi temporal Sentinel 1A satellite data with VV and VH polarization at 20&amp;thinsp;m spatial resolution was acquired between September 2016 and January 2017 at 12 days interval and processed using rule-based Parameterized classification in MAPscape-RICE software. Continuous monitoring for crop parameters and validation exercise was done for accuracy assessment. Spectral dB curve of rice was generated and the dB values ranged from &amp;minus;12.76 to &amp;minus;9.95 for VV and from &amp;minus;19.25 to &amp;minus;15.15 for VH polarization with an average primary variation of 1.3 and 2.5&amp;thinsp;dB respectively. Start of Season (SOS) map was derived from satellite data showing rice emergence dates for the cropping season. A total rice area of 106773&amp;thinsp;ha was estimated in Tiruvarur district using VV polarization with an overall accuracy of 79.5% and 0.59 kappa index, while in VH polarization, the rice area was estimated to be 91007&amp;thinsp;ha with 82.1% over all accuracy and 0.64 kappa index. The lesser accuracy in VV polarization was due to underestimate of direct seeded rice area and in VH polarization, it was due to underestimate in Transplanted rice area. The VV and VH rice area maps were then integrated to derive a VV-VH rice area map in MAPscape-RICE software and it recorded a total rice area of 124551 ha with an accuracy of 91.5% and 0.83-kappa index.</p>

scholarly journals INTEGRATING TIME-SERIES SAR DATA AND ORYZA CROP GROWTH MODEL IN RICE AREA MAPPING AND YIELD ESTIMATION FOR CROP INSURANCES

2019 ◽  
Vol XLII-3/W6 ◽  
pp. 239-243
Author(s):  
S. Pazhanivelan ◽  
K. P. Ragunath ◽  
N. S. Sudarmanian ◽  
R. Kumaraperumal ◽  
T. Setiyono ◽  
...  
Keyword(s):  
Time Series ◽  
Tamil Nadu ◽  
Crop Insurance ◽  
Crop Growth ◽  
Linear Diffusion ◽  
Temporal Features ◽  
Rice Area ◽  
Multi Temporal ◽  
Sar Data ◽  
Sar Imagery

<p><strong>Abstract.</strong> Lowland rice in tropical and subtropical regions can be detected precisely and its crop growth can be tracked effectively through Synthetic Aperture Radar (SAR) imagery, especially where cloud cover restricts the use of optical imagery. Parameterised classification with multi-temporal features derived from regularly acquired, C-band, VV and VH polarized Sentinel-1A SAR imagery was used for mapping rice area. A fully automated processing chain in MAPscape-Rice software was used to convert the multi-temporal SAR data into terrain-geocoded &amp;sigma;<sup>0</sup> values, which included strip mosaicking, co-registration of images acquired with the same observation geometry and mode, time-series speckle filtering, terrain geocoding, radiometric calibration and normalization. Further Anisotropic non-linear diffusion (ANLD) filtering was done to smoothen homogeneous targets, while enhancing the difference between neighbouring areas. Multi-Temporal Features viz., max, min, mean, max date, min date and span ratio were extracted from VV and VH polarizations to classify rice pixels. Rice detection was based on the analysis of temporal signature from SAR backscatter in relation to crop stages. About sixty images across four footprints covering 16 <i>samba</i> (<i>Rabi</i>) rice growing districts of Tamil Nadu, India were obtained between August 2017 and January 2018. In-season site visits were conducted across 280 monitoring locations in the footprints for classification purposes and more than 1665 field observations were made for accuracy assessment. A total rice area of 1.07 million ha was mapped with classification accuracy from 90.3 to 94.2 per cent with Kappa values ranging from 0.81 to 0.88. Using ORYZA2000, a weather driven process based crop growth simulation model developed by IRRI, yield estimates were made by integrating remote sensing products viz., seasonal rice area, start of season and backscatter time series. By generating average backscatter for each time series and dB stack for each SoS, LAI values were estimated. The model has generated rice yield estimate for each hectare which were aggregated at administrative boundary level and compared against CCE yield. Yield Simulation accuracy of more than 86&amp;ndash;91% at district level and 82&amp;ndash;97% at block level from the study indicates the suitability of these products for policy decisions. SAR products and yield information were used to meet the requirements of PMFBY crop insurance scheme in Tamil Nadu and helped in identifying or invoking prevented/failed sowing in 529 villages and total crop failure in 821 villages. In total 303703 farmers were benefitted by this technology in getting payouts of INR 9.94 billion through crop insurance. The satellite technology as an operational service has helped in getting quicker payouts.</p>

scholarly journals Rice Crop Monitoring and Yield Estimation Through Cosmo Skymed and TerraSAR-X: A SAR-Based Experience in India

2015 ◽  
Vol XL-7/W3 ◽  
pp. 85-92 ◽  
Author(s):  
S. Pazhanivelan ◽  
P. Kannan ◽  
P. Christy Nirmala Mary ◽  
E. Subramanian ◽  
S. Jeyaraman ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Food Security ◽  
Accuracy Assessment ◽  
Rice Crop ◽  
Policy Decisions ◽  
Simulation Accuracy ◽  
Sustainable Solutions ◽  
Rice Area ◽  
Sar Imagery ◽  
Crop Parameters

Rice is the most important cereal crop governing food security in Asia. Reliable and regular information on the area under rice production is the basis of policy decisions related to imports, exports and prices which directly affect food security. Recent and planned launches of SAR sensors coupled with automated processing can provide sustainable solutions to the challenges on mapping and monitoring rice systems. High resolution (3m) Synthetic Aperture Radar (SAR) imageries were used to map and monitor rice growing areas in selected three sites in TamilNadu, India to determine rice cropping extent, track rice growth and estimate yields. A simple, robust, rule-based classification for mapping rice area with multi-temporal, X-band, HH polarized SAR imagery from COSMO Skymed and TerraSAR X and site specific parameters were used. The robustness of the approach is demonstrated on a very large dataset involving 30 images across 3 footprints obtained during 2013-14. A total of 318 in-season site visits were conducted across 60 monitoring locations for rice classification and 432 field observations were made for accuracy assessment. Rice area and Start of Season (SoS) maps were generated with classification accuracies ranging from 87- 92 per cent. Using ORYZA2000, a weather driven process based crop growth simulation model; yield estimates were made with the inclusion of rice crop parameters derived from the remote sensing products viz., seasonal rice area, SoS and backscatter time series. Yield Simulation accuracy levels of 87 per cent at district level and 85- 96 per cent at block level demonstrated the suitability of remote sensing products for policy decisions ensuring food security and reducing vulnerability of farmers in India.

scholarly journals MULTI-TEMPORAL FEATURE EXTRACTION FOR PRECISE MAIZE AREA MAPPING USING TIME-SERIES SENTINEL 1A SAR DATA

2019 ◽  
Vol XLII-3/W6 ◽  
pp. 169-173
Author(s):  
M. Venkatesan ◽  
S. Pazhanivelan ◽  
N. S. Sudarmanian
Keyword(s):  
Time Series ◽  
Tamil Nadu ◽  
Growth Period ◽  
Maturity Stage ◽  
Kappa Score ◽  
Maximum Increase ◽  
Maize Crop ◽  
Temporal Features ◽  
Multi Temporal ◽  
Sar Data

<p><strong>Abstract.</strong> A research study was conducted to map maize area in Ariyalur and Perambalur districts of Tamil Nadu, India using multi-temporal features extracted from time-series Sentinel 1A SAR data. Multi-temporal Sentinel 1A GRD data at VV and VH polarizations and SLC products were acquired for the study area at 12 days interval and processed using MAPscape-RICE software. Multi-temporal Sentinel 1A data was used to identify the backscattering dB curve of maize crop. Analysis of temporal signatures of the crop showed minimum values at sowing period and maximum during the tasseling stage, which decreased during maturity stage of the crop. The maximum increase in the signature was observed during seedling to vegetative growth period. The signature derived from dB values for maize crop expressed a significant temporal behavior with the range of &amp;minus;21.26 to &amp;minus;13.18 in VH polarization and &amp;minus;14.05 to &amp;minus;6.54 in VV polarization. Considering the accuracy of SAR data to phenological variations of maize growing period, Multi-Temporal Features were extracted from multi-temporal dB images of VV and VH polarization and coherence images. Multi-Temporal Features viz., max, min, mean, max date, min date and span ratio were extracted from VV and VH polarizations of Sentinel 1A GRD and SLC data to classify maize pixels in the study area using parameterized classification approach. The overall classification accuracy was 91 percent with the kappa score of 0.82.</p>

Remote Sensing based Methane Emission from Rice Fields in Tiruchirapalli District

2017 ◽  
Vol 104 (4 - 6) ◽  
Author(s):  
Sudarmanian N S ◽  
◽  
Pazhanivelan S ◽  
Ragunath K P ◽  
Keyword(s):  
Satellite Data ◽  
Methane Emission ◽  
Regional Scale ◽  
Growth Period ◽  
Rice Fields ◽  
Anthropogenic Source ◽  
Diverse Range ◽  
Precise Estimation ◽  
Rice Area ◽  
Multi Temporal

Rice cultivation has been recognized as one of the major anthropogenic source for CH4 emissions. Methane emission from rice fields is a microbial mediated anaerobic activity, mainly favoured by the flooded condition. SAR-based operational mapping of rice crop across a diverse range of environments is possible with the increasing availability of multi-temporal SAR satellite data. Precise estimation of methane emission from rice fields at regional scale depends on accurate assessment of rice area and the corresponding time of flooding in those fields with IPCC emission factor. Start of Season (SoS) map was derived from satellite data showing rice emergence dates in Tiruchirapalli district recording 87 to 121 days of flooding during rice growth period. The rate of methane emission based on IPCC factor ranged from 37.4 to 45.74 kg/ha for a period of 87 to 121 days of flooding. The total methane emission from Tiruchirapalli district was 1.57Gg during Samba season 2015-16.

scholarly journals Sentinel 1A SAR Backscattering Signature of Maize and Cotton Crops

2017 ◽  
Vol 104 (.1-.4) ◽  
Author(s):  
Kumaraperumal R ◽  
◽  
Shama M ◽  
Balaji Kannan ◽  
Ragunath K P ◽  
...  
Keyword(s):  
Tamil Nadu ◽  
Ground Truth ◽  
Backscattering Coefficient ◽  
Ground Truth Data ◽  
Crop Monitoring ◽  
Multi Temporal ◽  
The Mean ◽  
Sar Data ◽  
Remote Sensing Techniques ◽  
Crop Discrimination

Crop discrimination is a key issue for agricultural monitoring using remote sensing techniques. Synthetic Aperture Radar (SAR) data are advantageous for crop monitoring and classification because of their all-weather imaging capabilities. The multi-temporal Sentinel 1A SAR data was acquired from 08th August, 2015 to 23rd January, 2016 at 12 days interval covering the extent of Perambalur district of Tamil Nadu. Both the Vertical - Vertical (VV) and Vertical-Horizontal (VH) polarized data are compared. The ground truth data collection was performed for cotton and maize during the vegetative, flowering and harvesting stages. The temporal backscattering coefficient (σ0 ) for cotton and maize are extracted using the training datasets. The mean backscattering values for cotton during the entire cropping period ranges from -10.58 dB to -6.28 dB and -20.59 dB to -14.53 dB for VV and VH polarized data respectively, and for maize it ranges from -11.08 dB to -7.07 dB and -19.85 dB to -14.14 dB for VV and VH polarized data respectively.

Estimation of crop parameters using multi-temporal optical and radar polarimetric satellite data

2015 ◽  
Author(s):  
Julie Betbeder ◽  
Remy Fieuzal ◽  
Yannick Philippets ◽  
Laurent Ferro-Famil ◽  
Frederic Baup
Keyword(s):  
Satellite Data ◽  
Multi Temporal ◽  
Crop Parameters

scholarly journals A Scalable Machine Learning Pipeline for Paddy Rice Classification Using Multi-Temporal Sentinel Data

2021 ◽  
Vol 13 (9) ◽  
pp. 1769
Author(s):  
Vasileios Sitokonstantinou ◽  
Alkiviadis Koukos ◽  
Thanassis Drivas ◽  
Charalampos Kontoes ◽  
Ioannis Papoutsis ◽  
...  
Keyword(s):  
Machine Learning ◽  
Satellite Data ◽  
High Performance ◽  
Large Scale ◽  
Paddy Rice ◽  
Machine Learning Algorithms ◽  
Classification Model ◽  
Supervised Machine Learning ◽  
Rice Area ◽  
Multi Temporal

The demand for rice production in Asia is expected to increase by 70% in the next 30 years, which makes evident the need for a balanced productivity and effective food security management at a national and continental level. Consequently, the timely and accurate mapping of paddy rice extent and its productivity assessment is of utmost significance. In turn, this requires continuous area monitoring and large scale mapping, at the parcel level, through the processing of big satellite data of high spatial resolution. This work designs and implements a paddy rice mapping pipeline in South Korea that is based on a time-series of Sentinel-1 and Sentinel-2 data for the year of 2018. There are two challenges that we address; the first one is the ability of our model to manage big satellite data and scale for a nationwide application. The second one is the algorithm’s capacity to cope with scarce labeled data to train supervised machine learning algorithms. Specifically, we implement an approach that combines unsupervised and supervised learning. First, we generate pseudo-labels for rice classification from a single site (Seosan-Dangjin) by using a dynamic k-means clustering approach. The pseudo-labels are then used to train a Random Forest (RF) classifier that is fine-tuned to generalize in two other sites (Haenam and Cheorwon). The optimized model was then tested against 40 labeled plots, evenly distributed across the country. The paddy rice mapping pipeline is scalable as it has been deployed in a High Performance Data Analytics (HPDA) environment using distributed implementations for both k-means and RF classifiers. When tested across the country, our model provided an overall accuracy of 96.69% and a kappa coefficient 0.87. Even more, the accurate paddy rice area mapping was returned early in the year (late July), which is key for timely decision-making. Finally, the performance of the generalized paddy rice classification model, when applied in the sites of Haenam and Cheorwon, was compared to the performance of two equivalent models that were trained with locally sampled labels. The results were comparable and highlighted the success of the model’s generalization and its applicability to other regions.

scholarly journals RICE (KHARIF) PRODUCTION ESTIMATION USING SAR DATA OF DIFFERENT SATELLITES AND YIELD MODELS: A COMPARATIVE ANALYSIS OF THE ESTIMATES GENERATED UNDER FASAL PROJECT

2019 ◽  
Vol XLII-3/W6 ◽  
pp. 99-107
Author(s):  
V. Jain ◽  
S. Saxena ◽  
S. Dubey ◽  
K. Choudhary ◽  
S. Sehgal ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Accuracy Assessment ◽  
Monsoon Season ◽  
Ground Truth ◽  
Coefficient Of Determination ◽  
Optical Remote Sensing ◽  
Acreage Estimation ◽  
Statistical Measures ◽  
Multi Temporal ◽  
Sar Data

<p><strong>Abstract.</strong> Rice is the most important food crop of India. Majority of Rice is sown in kharif season in the country. This is monsoon season for the country where cloud cover poses a major problem for optical remote sensing. Therefore, for these states rice acreage estimation is being done using Synthetic Aperture Radar (SAR) data operationally in India since 1998. A case study is presented in this paper for analysis of past 6 years’ (2012&amp;ndash;13 to 2017&amp;ndash;18) estimations. Multi temporal Radarsat-2 (HH), RISAT-1 ScanSAR (HH) and Sentinel-1 (VV) data was used in years 2012, 2013&amp;ndash;2016, and 2017, respectively for paddy identification. Hierarchal Decision Rule based classification (HDRC) approach was used to identify rice areas under sample segments. Extensive ground truth collected by state remote sensing departments and agriculture departments was utilized in setting the limits of HDRC models and accuracy assessment. Yield was estimated using weather based and remote sensing-based models. Area, production and yield estimates were made and compared with those given by DES. RMSE and R<sup>2</sup> were used as statistical measures to assess the accuracy of results. The RMSE % ranged from 2.3 to 4.3; 0.84 to 1.35; 0.24 to 0.27 for area, production and yield respectively. The coefficient of determination (R<sup>2</sup>) ranged from 0.62 to 0.92; 0.75 to 0.91; 0.5 to 0.83 for area, production and yield respectively. The study showed that use of multi temporal SAR data (both HH and VV) is quite useful for paddy acreage estimation, especially during monsoon.</p>

scholarly journals AREA ESTIMATION OF COTTON AND MAIZE CROPS IN PERAMBALUR DISTRICT OF TAMIL NADU USING MULTI DATE SENTINEL-1A SAR DATA &amp; OPTICAL DATA

2019 ◽  
Vol XLII-3/W6 ◽  
pp. 137-140
Author(s):  
K. Ramalingam ◽  
A. B. Ramathilagam ◽  
P. Murugesan
Keyword(s):  
Tamil Nadu ◽  
Unsupervised Classification ◽  
Optical Data ◽  
Landsat 8 ◽  
Maize Crop ◽  
Classification Approach ◽  
Area Estimation ◽  
Crop Calendar ◽  
Multi Temporal ◽  
Sar Data

<p><strong>Abstract.</strong> This study was carried out to estimate the area of cotton and maize crops in Permabalur district of Tamil Nadu using microwave and optical data. Permabalur was selected as the study area, as it is the largest cotton and maize producing district in Tamil Nadu. The multi-temporal Sentinel-1A SAR data was acquired from 09th July, 2016 to 17th January, 2017 as it coincides with the crop calendar of these crops. Both the Vertical-Vertical (VV) and Vertical-Horizontal (VH) polarized data were compared. The cloud free Landsat 8 data acquired on 7th October 2016 was fused with the Vertical–Vertical (VV) and Vertical-Horizontal (VH) polarized data of 13th October and classified. Unsupervised classification approach was adopted to classify the cotton and maize pixels. The highest accuracy of 72.73% and 76.24% were achieved in VV polarization + Landsat 8 data and VH polarization + Landsat 8 data respectively. The cotton and maize areas were estimated to be 20,218&amp;thinsp;ha and 28,032&amp;thinsp;ha respectively. It is also evident that VH polarization fused with optical data is better in discriminating the cotton and maize crop than VV polarization fused with optical data.</p>

Sign in / Sign up

Export Citation Format

Share Document