University of Twente Student Theses
Estimation of Transplanting and Harvest Dates of Rice Crops in the Philippines Using Sentinel-1 Data
Cauba, A. (2023) Estimation of Transplanting and Harvest Dates of Rice Crops in the Philippines Using Sentinel-1 Data.
|
PDF
4MB |
| Abstract: | Rice is an important staple crop in the Philippines; hence, accurate estimation of its transplanting and harvest dates is essential for efficient crop management and resource allocation. Remote sensing data, such as Sentinel-1 Synthetic Aperture Radar (SAR) data, offer valuable information for monitoring agricultural activities and predicting crop phenology. This study focused on utilizing Sentinel-1 SAR data to estimate the transplanting and harvest dates of rice crops in two distinct seasons, namely the dry and the wet seasons, in the provinces of Agusan del Sur, Cagayan, and Leyte. Transplanting and harvest dates of rice fields were obtained through farmer interviews conducted by the International Rice Research Institute (IRRI). A total of 99 rice fields were further considered (19 in Agusan del Sur, 38 in Cagayan and 42 in Leyte) in this study. Time series data from Sentinel-1A (obtained from IRRI) and Sentinel-1B (extracted from Google Earth Engine) were merged, and the mean backscatter coefficients in VV, VH, and VH/VV polarizations were extracted at the rice field level to create time series curves. The locally weighted scatterplot smoothing (LOWESS) method was applied to smoothen the time series data, and the R2 values were used to assess the goodness of fit between the original and smoothed curves. Periodogram analysis and the Breusch-Godfrey test were employed to identify repetitive patterns and their statistical significance. Local extrema and corresponding dates were identified to indicate transplanting and harvest dates. The detected dates were then compared with those from the field survey data provided by the IRRI. The evaluation of the goodness of fit revealed the variability of the time series data. VV polarization data during the wet season exhibited the lowest R2 values (median R2 < 0.4), with Agusan del Sur and Leyte provinces demonstrating higher variability than the Cagayan province. Conversely, VH data consistently displayed the highest R2 values across all three provinces (median R2 > 0.59), indicating stronger linear relationships between the original and LOWESS-smoothed time series curves. Around the transplanting period, the backscatter coefficients decreased, resulting in minimal backscatter. We observed some discrepancies between the transplanting dates reported by farmers and our detected dates. During the dry season, the lowest root mean squared differences (RMSD) for transplanting dates were 9.2, 16, and 13.6 days and during the wet season, were 14, 29, and 27 days in Agusan del Sur, Cagayan, and Leyte, respectively. When estimating the harvest dates, VH and VH/VV polarization consistently corresponded to the local maximum immediately after the first local minimum. In contrast, VV polarization dates were associated with either the first or second local maximum after the first local minimum. During the dry season, the lowest RMSD for harvest dates were 17.5, 16, and 16 days and during the wet season were 8, 14, and 22 days in Agusan del Sur, Cagayan, and Leyte, respectively. In a nutshell, our results revealed the potential of VH and VV polarizations for estimating transplanting and harvest dates during the dry season, while VH/VV polarization showed promising results for estimating these dates mainly during the wet season. This research has significant implications for the agricultural sector, providing opportunities to optimize crop management practices, allocate resources effectively, and support sustainable rice production. |
| Item Type: | Essay (Master) |
| Faculty: | ITC: Faculty of Geo-information Science and Earth Observation |
| Programme: | Geoinformation Science and Earth Observation MSc (75014) |
| Link to this item: | https://purl.utwente.nl/essays/95167 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page