Tracking penguin populations has never been easier, thanks to a new SOOS-endorsed and NASA-funded decision support tool called the Mapping Application for Penguin Populations and Projected Dynamics (MAPPPD). MAPPPD (pronounced “mapped”) can be found at www.penguinmap.com and represents a collaboration between Dr. Heather Lynch at Stony Brook University, Dr. Mathew Schwaller at NASA Goddard, and the Washington, DC-based research and education foundation Oceanites, Inc.
The goals of the project are to:
(1) automate the interpretation of high- and medium-resolution satellite imagery for estimating the abundance and distribution of penguins in Antarctica; and
(2) integrate these data into scalable abundance estimates at any user-defined spatial scale.
There is also a strong citizen science component to MAPPPD, and the website contains all the tools needed to get citizen scientists searching Google Earth for new or previously undiscovered penguin colonies.
Figure 1: Screenshot of the MAPPPD map showing penguin populations.
Increased availability of satellite imagery for the Antarctic provides new opportunities for monitoring Antarctic biology, but poses unique challenges for data synthesis and interpretation. MAPPPD includes data on the abundance and distribution of the four penguin species with substantial Antarctic breeding populations: the emperor penguin (Aptenodytes forsteri), the Adélie penguin (Pygoscelis adeliae), the gentoo penguin (P. papua), and the chinstrap penguin (P. antarcticus). These data derive from direct ground surveys or aerial photographic surveys. However, MAPPPD will also include near real-time streaming data as obtained from satellite imagery, including both medium resolution sensors (e.g., Landsat) and high-resolution commercial sensors such as Quickbird and Worldview.
The easy-to-use interface ensures that data are accessible and easily used by both the scientific and non-scientific Antarctic communities. Users can search colonies by name, using an interactive map, or by CCAMLR sub-area. Users can also upload shapefiles representing regions of particular interest, such as in the design of protected areas. The expected result is a flexible all-purpose tool for managing, interpreting, and visualizing the abundance and distribution of penguins in Antarctica.
Figure 2: Nesting Adelie penguin (Photo by Heather Lynch)
MAPPPD creates, organizes, and displays data on the abundance and distribution of penguin species at spatial and temporal scales commensurate with other remote observing systems in Antarctica, allowing for smooth interoperability across the physical and biological realms. MAPPPD provides a regional and continental-scale context for additional observations of Antarctic biology, such as those stemming from animal-borne instruments or opportunistic records of species occurrence. It thus provides a key dataset for helping to understand the impact of global change on Southern Ocean ecosystems. Moreover, the use of remote sensing to track the abundance and distribution of penguins expands the options for understanding Antarctic biology; includes the integration of data coming from direct surveys; and naturally complements the work already underway with animal sensors.
One of the most unusual aspects of MAPPPD will be its tools for model comparison. Models displayed in MAPPPD will be available through a public github repository, which can then be cloned by other scientists interested in model development. Community-developed models will then be included in MAPPPD for display and download, and new data each year will be used to evaluate the predictive performance of the models in MAPPPD. Incorporating a collection of models permits the generation of ensemble model forecasts and facilitates the study of “model uncertainty”. Tools for models download, upload, and comparison are due to be completed in 2017. In the meantime, the database search tools are proving popular, and MAPPPD is already helping to clarify the often-confusing literature on historical penguin censuses.