Jonsson and Eklundh in TIMESATsoftware packagewas used to fit the curve of time series data following second order polynomial and to extract the following phenological metrics 19, 20, 21, 22. - Start of growing season (SOS): This is defined as the dates of leaf unfolding, this study considered SOS as a date when NDVI increases to 5% of the amplitude during the beginning period of the NDVI time series. For the series of NDVI in a given NDVImax year, is the maximum of NDVI and NDVImin is minimum of NDVI in the first half of the year. NDVIstart was calculated as follows; - End of the season (EOS): This is defined as the dates leaf discoloration and leaf fall at the end of season. This study considered EOS as a date when NDVI decreases to 5% of the amplitude during the ending period of the NDVI time series. NDVIend was calculated as follows; - Length of the growing season (LOS) is the number of days of growth from the start to the end of this season. The results of 5% changing in NDVI value from minimum for SOS and EOS agreed to the transition date of LAI index and NLR, therefore this threshold has been applied for extracting seasonality data such as start, end of the season and the length of the season in this study. 2) Relationship between NDVI and Climate Factors: Daily data at the meteorological station in Lampang Province were aggregated to an 8 day period by calculating average for temperature and accumulated precipitation. Base on the distribution of data, the non- parametric Spearman was applied to check the correlation coefficient between NDVI and climate factors 23. III. RESULTS AND DISCUSSION A. NDVI Time Series of Teak Plantation The variation of NDVI, precipitation, minimum and maximum temperature at Teak plantation are shown in Fig.2, 3 and 4 respectively.Rainy season starts between March to May, the same time as the teak leaf starts to bud reflecting the gradual increase of NDVI values. Rainy season ends between October and November, corresponding to the decrease of NDVI values in this period of time (Fig.2). During April which is the hottest month, NDVI value was lowest, indicating the leafless period of teak plantation. NDVI values increased from 0.35-0.40 to 0.75-0.80 during May to October, and then it decreased in November and December (Fig.3, 4). Growing season ended in January, February or March in different areas. The linear correlation between NDVI with LAI and NLR was found, confirming the cloud free data set. It was found that NDVI was significantly correlated with LAI (r=0.81, p<0.0001), and NLR (r=0.77, p<0.0001) (Fig.6). This correlation is consistent with other previous reports. For example, the relationship between NDVI and field measured LAI in deciduous forests were reported by Madugundu et al. 24. In their study, NDVI was strongly correlated with LAI (r2=0.68, p≤ 0.05) but showed evidence of saturation above a biomass of 100 g/m2 and an LAI of 2 m2/m2 25. Potithep et. al. 23 has also found that NDVI changes in relation to the seasonal variations of LAI 26. These results indicate that the preprocessing NDVI derived from satellite is sufficient to capture the teak forest phenology. Proceedings of the International Conference on Climate Change, Biodiversity and Ecosystem Services for the 277 Sustainable Development Goals (SDGs): Policy and Practice 27-29 June 2016, Cha-am, Phetchaburi, Thailand
Proceedings of International Conference on Climate Change, Biodiversity and Ecosystem Services for the Sustainable Development Goals : Policy and Practice 27-29 June 2016 at the Sirindhorn International Environmental Park, Cha-am, Phetchaburi, Thailand
To see the actual publication please follow the link above