Satellite based phenology study in Southeast Asia

Phenology is the study of the response of living organisms to seasonal and climatic changes to the environment in which they live. Seasonal changes include 'variations in the duration of sunlight, precipitation, temperature and other life- controlling factors. Phenology is generally accepted as including not only the timing of recurring biological events but also their causes, especially with regard to meteorological phenomena (Leith 1974). Vegetation phenology is an important variable in a wide variety of Earth and atmospheric science applications.

The study of phytophenology is important in its own right for the insight it gives into the temporal organization, evolution and functioning of ecosystems. Thus, phytophenological studies are concerned with the influence of seasonally varying environmental conditions such as day-length, air temperature and water availability on the timing of plant development stages.

One of the outstanding merits of remote sensing procedures in the investigation of phenology relates to the spatially comprehensive overview of vegetation that is provided. Seasonal changes in the greenness of vegetation are described in remotely sensed data as changes in the Normalized Difference Vegetation Index (NDVI) throughout the year.

Southeast Asia was chosen in this study, because: (1) Their size and shape are compatible with the 1 km x 1 km spatial resolution of NOANAVHRR satellite data, (2) Southeast Asia as a definable world region and have provided a framework for the comparative study of its components, and (3) Southeast Asia region was located by superimposing a geographic database built at the laboratory on AVHRR imagery. This region covers locations from Northern Australia in the South (15° S) up to the Southern China (25° N) and from Bangladesh (90° E) to Irian Jaya (140° E)

There are four primary data sets used in this study, namely: (1) AVHRR-NOAA satellite data for May 1997 to April 1998 period, (2) Digital Elevation Model (DEM), (3) Olson World Ecosystem (land cover), and (4) Hydrometeorological data respectively.

By means of the Remote Sensing and Geographic Information System (GIS) techniques, seven different phenological metrics (its distribution and variability) were defined for land cover types analyzed in this study: (1) NDVI threshold, (2) total of greenness onset, (3) time of greenness onset, (4) duration of greenness, (5) time of maximum NDVI, (6) value of maximum NDVI, and (7) time-integrated NDVI. Each land cover types pointed out the distinct characteristics of phenological events.

Forests land cover types are commonly green throughout the year (as growing season period). However, this event was not went through by all the conifer forest types, part of the deciduous mixed forest types. In general, forest types have the average of maximum NDVI value are higher the other types. Moreover, the rate of carbon consumption for forest types is also usually higher, while there is exception in deciduous mixed forest.

Good phenomena in the phenological events were illustrated by cropland /irrigation, grassland and interrupted woods types. These types got into twice or three times growing season period or planting season in a year, in which the greenness month is longer than harvest time.

All the land cover types analyzed in the study area start the beginning of greenness onset on the first May 1997. Apart from that, planting season over Southeast Asia in the low latitude (10°N to 10°S) is generally longer than else. The whole land covers have the ability of consuming carbon, even if it is various.

Effect of hydrometeorological events in promoting vegetation growth over Southeast Asia for May 1997 to April1998 was only undertaken by annual rainfall. The result pointed out that the NDVI relates to spatial changing patterns of annual rainfall, in which there is exponential relationship between NDVI and annual rainfall. This exponential relationship means that the NDVI relates to changing patterns of annual rainfall only in the dry regime, with a rainfall regime of less than approximately 1500 mm per year and/or 125 mm per month.