Abstract

Abstract. Knowing where and when rice is grown is essential for planning and decision-making in relation to food security, as well as in research wherein crop area and calendar are important inputs in crop production simulations, assessment of biotic and abiotic stresses, and analysis of the effect of climate change on crop production, among others. Remote sensing allows for efficient mapping and characterization of rice areas. In this study, we derived the rice planting window in all rice growing regions in the Philippines from 2016 to 2018 using multi-temporal Synthetic Aperture Radar (SAR), specifically TerraSAR-X and Sentinel-1. Using a rule-based method, rice area and Start of Season (SoS) were mapped based on the unique backscatter behaviour of rice corresponding to the initial deliberate agronomic flooding followed by rapid biomass increase. We defined the planting window per year and semester as the 15th and 85th percentile and the peak of planting as the dominant planting date. The accuracy of the rice map was 93% and the SoS was strongly correlated with the actual planting dates reported by farmers (R2 = 0.71) based on 482 ground observations in the Philippines in 2018 Semester 1. From this analysis, the planting window in the Philippines for the Semester 2 (wet season) is April-August (peak in June-July), and for Semester 1 (dry season) is September-February (peak in November-December) with large differences across regions. In majority of the regions, the planting window spans more than 100 days, which can have implications on incidence of pests and diseases.

Highlights

  • Rice remains the top agricultural commodity in the Philippines accounting for 93% of household consumption in 2015/2016 (PSA, 2017)

  • Start of season was detected based on the value of reliability coefficient (RC) computed after satisfying the following conditions: (1) Synthetic Aperture Radar (SAR) backscatter value (VV and VH polarization), (2) increase in backscatter observed after the start of season (SoS), (3) correspondence of SoS date in VV and VH polarizations, and (4) value of local incidence angle (LIA) in overlapping areas

  • In Semester 1 the peak of SoS was detected in November/December (Visayas and Mindanao) and December/January (Luzon) while the range started from September and the latest was March across years

Read more

Summary

Introduction

Rice remains the top agricultural commodity in the Philippines accounting for 93% of household consumption in 2015/2016 (PSA, 2017). Recent studies show that characteristics of crops such as the start of season (SoS), phenological stages, and even standing water beneath vegetation can be detected using multi-temporal Synthetic Aperture Radar (SAR) (Martinis, Rieke, 2015; Torbick et al, 2017; Bargiel, 2017). Spaceborne SAR can accommodate wider area coverage at a higher resolution (Qiu et al, 2013). Mapping of rice areas and detection of crop planting dates using SAR is a time efficient and cost effective means of obtaining up-to-date information (More et al, 2016)

Objectives
Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call