MODIS Derived Primary Productivity: Seasonal Hotspots

Brief description

The dataset indicates the seasonal primary productivity hotspots of ocean surface waters. They are derived from MODIS (aqua) images using NASA's SeaDAS image processing software. The monthly chlorophyll a images between July 2002 and August 2014 are used to identify the primary productivity hotspots of the four austral seasons: winter (June, July, and August), spring (September, October and November), summer (December, January and February) and autumn (March, April and May). The extent of the dataset covers the entire Australian EEZ and surrounding waters (including the southern ocean). The value (between 0 and 1.0) of the dataset represents the likelihood of the location being a primary productivity hotspot.

This research is supported by the National Environmental Science Program (NESP) Marine Biodiversity Hub through Project D1.

Lineage

Maintenance and Update Frequency: asNeeded

Statement: The daily MODIS Aqua Chlorophyll a images processed to Level 2 were obtained from the Integrated Marine Observing System (IMOS; http://imos.org.au/) remote sensing facility. The Chlorophyll a algorithm used is the semi-analytical GSM algorithm (Garver and Siegel 1997; Maritorena et al. 2002). The monthly Chlorophyll a images were generated by mosaicking the daily Chlorophyll a images, where the overlapping locations have average values of the overlapping cells. The topographic position index (TPI) (Weiss 2001) images were then calculated from the monthly Chlorophyll a images. A positive and large TPI value indicates a likely primary productivity hotspot location. Next , the TPI images were converted into the hotspot likelihood images by using the following rules:
1. If the TPI value is greater than 0.5 time of the TPI image's spatial standard deviation (STD) then the likelihood equals to 1.0.
2. If the TPI value is smaller than 0.1 time of TPI image's spatial standard deviation (STD) then the likelihood equals to 0.
3. If the TPI value is greater than 0.1 time but smaller than 0.5 time of TPI image's spatial standard deviation (STD) then the likelihood is calculated using this linear equation: (TPI - 0.1*STD)/0.4*STD.
It should be noted that the TPI images have variable spatial standard deviations.

The seasonal means were then calculated from the monthly primary productivity likelihood images of the corresponding seasons between July 2002 and August 2014. For example, the winter mean was obtained from all June, July and August images. The spring mean was obtained from all September, October and November images. The summer mean was obtained from all December, January and February images. The autumn mean was obtained from all March, April and May images.

The Great Barrier Reef and a number of offshore reefs have been masked out due to the unreliable MODIS Chlorophyll a values known to occur at these shallow water areas.