Institute of Environmental Science and Meteorology (IESM)
University of the Philippines
Diliman, Quezon City, Metro Manila, The Philippines Abstract of Talk:
The changes in global climate patterns pose considerable risks to biologically diverse and agriculturally dependent countries such as the Philippines. With a rapid forest loss of about 40,000 hectares (ha) per year during the period 1500 to 2004, the remaining forests in the Philippines cover 7.2 Million ha or only 24.3% of the country’s total land area, less than 3% of which is covered by primary forests. Preservation of these pristine areas is highly critical to the biodiversity associated with it. The Philippines is one of the few mega-diverse countries in the world with more than 65% of the species found nowhere else. Realizing the importance of keeping this biodiversity, laws have been passed to prevent the further decline of the remaining forest and to initiate reforestation efforts. To ensure that these efforts are effectively implemented, satellite data are needed to adequately monitor the changes in vegetation cover. At the same time, satellite data can be used to effectively manage other natural resources including agricultural areas and grasslands that usually replace old forested areas. Philippine agriculture represents 1/5 of the total economy (18% of GDP) and generates 1/3 of the country’s total employment and provides food and livelihood to its people. For optimum productivity, surface temperature, soil moisture, rainfall and the state of vegetation need to be monitored, all of which can be done efficiently with satellite data.
Philippine vegetation and its seasonal and interannual variability from 2000 to the present are studied using the Normalized Difference Vegetation Indexes (NDVI) derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) products. The use of the relatively high 250 m resolution data in the visible and near infrared channels has been found to be very useful in quantifying general locations of forest areas and agricultural regions and how they change from one season to another. The key dataset used is the 16-day composite NDVIs. A special procedure for cloud masking was implemented because using the strictly confident cloud-free version of the data provided only 15% spatial coverage that is not suitable for time series analysis, whereas, using near confident clear skies version provided 60% coverage. With the technique that ensures elimination of suspicious cloud covered data, the resulting data set is shown to be spatially and temporally coherent and have values that are physically viable as dictated by location and season. Validation of the final product was made and the NDVI values show good agreement with high-resolution NDVI data such as those derived from LANDSAT. Satellite data were also used to monitor other geophysical parameter; namely, surface temperature data, which were derived from MODIS and the rainfall data, which were derived from the Tropical Rainfall Measuring Mission (TRMM).
Analyses of the vegetation dataset show large seasonal variability in lowlands or the generally agricultural areas, while moderate seasonality is observed in the high lands where the vegetation consists of primary forests. NDVI maxima are observed during the rainy months (June to November) with moderate interannual variability while NDVI minima are recorded in the dry/hot months (March to May) with much stronger interannual variability. The interannual variability is reflected in most part of the country with the exception of the southern region (i.e., Mindanao) where the NDVI is relatively high and stable. The record shows enhanced vegetation during the dry/hot seasons of 2000, 2001, 2008 and 2009 and anomalously low vegetation in 2003, 2005, 2007 and 2010. The lowest value of NDVI from 2000 to 2010 was observed in the summer of 2010, with a decline of 4.7% from the previous year compared to an average fluctuation of 1.8%. This event coincided with an anomalously warm and dry year in 2010. The abnormally low NDVI in 2010 was studied in the context of the potential impact of global warming on vegetation in the Philippines. The lack of rainfall during the period also implies that drought conditions were possible. During the hot/dry season of this year, the most affected regions were the lowland areas that include agricultural regions. In Luzon, the most seriously affected vegetation areas are those in the Cagayan Valley and the western part of Central Luzon. In the Visayas, NDVI was extremely low in Negros and Iloilo. These same regions have the highest anomalies in surface temperature and should be regarded as the regions that are most vulnerable to greenhouse warming. Through thresholding techniques, the range of NDVI values that are associated with perennial vegetation were established. This enabled quantitative estimates of the extent of remaining forest at high elevation areas, which was found to be declining at the rate of about 1% per decade (2000 to 2010). Future studies will include in depth analysis of the loss of vegetation that may be associated with deforestation, fire and other disturbances.