Paddy-Rice Phenology Classification Based on Machine-Learning Methods Using Multitemporal Co-Polar X-Band SAR Images

Küçük, Çağlar; Taşkın, Gülşen; Erten, Esra

doi:10.1109/jstars.2016.2547843

Cited by 71 publications

(20 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Support Vector Machine (SVM) classification with the Kernel function [72][73][74] was used for LULC mapping. The RBF parameters were optimal based on the classification results of the polarimetric SAR data [73][74][75][76][77][78]. The optimal parameters were grid search and cross-validation, which were introduced by previous studies [79,80].…”

Section: Data Analysis and Accuracy Assessmentmentioning

confidence: 99%

Monitoring and Mapping of Rice Cropping Pattern in Flooding Area in the Vietnamese Mekong Delta Using Sentinel-1A Data: A Case of An Giang Province

Minh

Avtar

Mohan

et al. 2019

IJGI

View full text Add to dashboard Cite

Cropping intensity is one of the most important decisions made independently by farmers in Vietnam. It is a crucial variable of various economic and process-based models. Rice is grown under irrigated triple- and double-rice cropping systems and a rainfed single-rice cropping system in the Vietnamese Mekong Delta (VMD). These rice cropping systems are adopted according to the geographical location and water infrastructure. However, little work has been done to map triple-cropping of rice using Sentinel-1 along with the effects of water infrastructure on the rice cropping intensity decision. This study is focused on monitoring rice cropping patterns in the An Giang province of the VMD from March 2017 to March 2018. The fieldwork was carried out on the dates close to the Sentinel-1A acquisition. The results of dual-polarized (VV and VH) Sentinel-1A data show a strong correlation with the spatial patterns of various rice growth stages and their association with the water infrastructure. The VH backscatter (σ°) is strongly correlated with the three rice growth stages, especially the reproductive stage when the backscatter is less affected by soil moisture and water in the rice fields. In all three cropping patterns, σ°VV and σ°VH show the highest value in the maturity stage, often appearing 10 to 12 days before the harvesting of the rice. A rice cropping pattern map was generated using the Support Vector Machine (SVM) classification of Sentinel-1A data. The overall accuracy of the classification was 80.7% with a 0.78 Kappa coefficient. Therefore, Sentinel-1A can be used to understand rice phenological changes as well as rice cropping systems using radar backscattering.

show abstract

Section: Data Analysis and Accuracy Assessmentmentioning

confidence: 99%

Monitoring and Mapping of Rice Cropping Pattern in Flooding Area in the Vietnamese Mekong Delta Using Sentinel-1A Data: A Case of An Giang Province

Minh

Avtar

Mohan

et al. 2019

IJGI

View full text Add to dashboard Cite

show abstract

“…Additionally, dense time series are necessary to understand the Synthetic Aperture Radar (SAR) signal behavior with regards to crops. From the first attempt to monitor rice crops [23], relevant results were found by combining different SAR sensors [24], incidence angles [25], different polarizations [26][27][28][29][30][31], and Interferometric SAR technique [32]. A data fusion approach was developed using a dynamical framework based on particle filter (PF).…”

Section: Introductionmentioning

confidence: 99%

Exploiting Time Series of Sentinel-1 and Sentinel-2 Imagery to Detect Meadow Phenology in Mountain Regions

et al. 2019

View full text Add to dashboard Cite

A synergic integration of Synthetic Aperture Radar (SAR) and optical time series offers an unprecedented opportunity in vegetation phenology monitoring for mountain agriculture management. In this paper, we performed a correlation analysis of radar signal to vegetation and soil conditions by using a time series of Sentinel-1 C-band dual-polarized (VV and VH) SAR images acquired in the South Tyrol region (Italy) from October 2014 to September 2016. Together with Sentinel-1 images, we exploited corresponding Sentinel-2 images and ground measurements. Results show that Sentinel-1 cross-polarized VH backscattering coefficients have a strong vegetation contribution and are well correlated with the Normalized Difference Vegetation Index (NDVI) values retrieved from optical sensors, thus allowing the extraction of meadow phenological phases. Particularly for the Start Of Season (SOS) at low altitudes, the mean difference in days between Sentinel-1 and ground sensors is compatible with the acquisition time of the SAR sensor. However, the results show a decrease in accuracy with increasing altitude. The same trend is observed for senescence. The main outcomes of our investigations in terms of inter-satellite comparison show that Sentinel-1 is less effective than Sentinel-2 in detecting the SOS. At the same time, Sentinel-1 is as robust as Sentinel-2 in defining mowing events. Our study shows that SAR-Optical data integration is a promising approach for phenology detection in mountain regions.

show abstract

“…However, the initial interest was likely targeted towards several different research objectives, without a specific rs+pheno focus, to such an extent that these terms gained a high relevance only in the last decade. The use of new sensors and specific optical and radar bands [84][85][86][87], their data fusion [55,88], as well as the application of yield prediction and crop simulation models [89][90][91] may be, hence, considered as a new challenge for the next decade of rs+pheno studies.…”

Section: Emerging Research Topicsmentioning

confidence: 99%

Text Mining in Remotely Sensed Phenology Studies: A Review on Research Development, Main Topics, and Emerging Issues

et al. 2019

View full text Add to dashboard Cite

As an interdisciplinary field of research, phenology is developing rapidly, and the contents of phenological research have become increasingly abundant. In addition, the potentiality of remote sensing technologies has largely contributed to the growth and complexity of this discipline, in terms of the scale of analysis, techniques of data processing, and a variety of topics. As a consequence, it is increasingly difficult for scientists to get a clear picture of remotely sensed phenology (rs+pheno) research. Bibliometric analysis is increasingly used for the study of a discipline and its conceptual dynamics. This review analyzed the last 40 years of publications in the rs+pheno field retrieved from the Scopus database; such publications were investigated by means of a text mining approach, both in terms of bibliographic and text data. Results demonstrated that rs+pheno research is exponentially growing through time; however, it is primarily considered a subset of remote sensing science rather than a branch of phenology. In this framework, in the last decade, agriculture is becoming more and more a standalone science in rs+pheno research, independently from other related topics, e.g., classification. On the contrary, forestry struggles to gain its thematic role in rs+pheno studies and remains strictly connected with climate change issues. Classification and mapping represent the major rs+pheno topic, together with the extraction and the analysis of phenological metrics, like the start of the growing season. To the contrary, forest ecophysiology, in terms of ecosystem respiration and net ecosystem exchange, results as the most relevant new topic, together with the use of the red edge band and SAR (Synthetic Aperture Radar) data in rs+pheno agricultural studies. Some niche emerging rs+pheno topics may be recognized in the ocean and arctic investigations linked to phytoplankton blooming and ice cover dynamics. The findings of this study might be applicable for planning and managing remotely sensed phenology research; scientists involved in such discipline might use this study as a reference to consider their research domain in a broader dynamical network.

show abstract

Paddy-Rice Phenology Classification Based on Machine-Learning Methods Using Multitemporal Co-Polar X-Band SAR Images

Cited by 71 publications

References 31 publications

Monitoring and Mapping of Rice Cropping Pattern in Flooding Area in the Vietnamese Mekong Delta Using Sentinel-1A Data: A Case of An Giang Province

Monitoring and Mapping of Rice Cropping Pattern in Flooding Area in the Vietnamese Mekong Delta Using Sentinel-1A Data: A Case of An Giang Province

Exploiting Time Series of Sentinel-1 and Sentinel-2 Imagery to Detect Meadow Phenology in Mountain Regions

Text Mining in Remotely Sensed Phenology Studies: A Review on Research Development, Main Topics, and Emerging Issues

Contact Info

Product

Resources

About