Page 64 - D:\Video IPSyoFS22\

Mapping of rice extent and growth stages across Peninsular Malaysia using sentinel-1 and 2
data

* 1
1
1 Fatchurrachman, Rudiyanto , Soh Norhidayah Che, Shah Ramisah Mohd
1

1 Program of Crop Science, Faculty of Fisheries and Food Science, Universiti Malaysia Terengganu,
Kuala Nerus 21030, Terengganu, Malaysia.

* Corresponding author: rudiyanto@umt.edu.my

Abstract:
Rice is the staple crop for more than half the world’s population. Nevertheless, high-resolution maps of
rice extent and its growth stages are lacking. Most maps produced by remote sensing technology only
provide the rice extent. However, rice in tropical regions is grown throughout the year with high
variations in planting dates and frequency. Thus, mapping rice growth stages could give more valuable
information instead of mapping the rice extent only. This study addressed this issue by developing a
phenology-based method. The hypothesis was that the k-means clustering method of Sentinel-1 and 2
time-series data could identify rice fields and growth stages, because flooding during transplanting stage
can be identified by Sentinel-1 VH backscatter; and changes in the canopy of rice fields during growth
stages (vegetative, generative and ripening phases) up to harvesting stage can be distinguished by
Sentinel-2 Normalized Difference Vegetation Index (NDVI). This study used the proposed method to
develop a rice field extent map and cropping calendars across Peninsular Malaysia (130,598 km2) on
the Google Earth Engine (GEE) platform. The Sentinel-1 and 2 monthly time series data from January
2019 to December 2020 were classified using k-means clustering to identify areas with similar
phenological patterns. This study resulted in high-accuracy 10-meter resolution maps of rice extent,
intensity and cropping calendars. Validation using very high-resolution street view images from Google
Earth showed that the produced map had an overall accuracy of 95.95%, with a kappa coefficient of
0.92. Furthermore, the predicted cropping calendars coincided well with the government’s agency data.
In conclusion, the proposed method is cost-effective and have capability to accurately map rice fields
extent and growth stages over large areas. The resulted data will be useful in measuring the
accomplishment of rice production self-sufficiency and the estimation of methane emissions from rice
cultivation.

Keywords: Paddy fields, Phenology, Sentinel-1, Sentinel-2, Google Earth Engine