Censusing Animal Populations from Space (CAPS)
Developing a cost-effective, remote sensing-based method for monitoring animal populations from space
Despite being the greatest consumers of krill in the Southern Ocean, our understanding of the status and trends in pack-ice seal populations and their relationship with key habitat characteristics, such as sea-ice, currently represents a major knowledge gap. Until now, it has been too logistically challenging and expensive to conduct regular pack-ice seal surveys at a spatial scale sufficient to assess their regional-scale abundance and distribution. As a result, pack-ice seals have been largely neglected, with the notable exception of a large international survey of pack-ice seals conducted between 1999-2001 (APIS I). Remote sensing based methods provide a cost-effective approach to monitoring seals that will enable these otherwise elusive data be made available. This working group contributes to the SOOS Science Priority Theme 6, impacts of global change on Southern Ocean ecosystems, by advancing the development of cost-effective monitoring approaches to assess changes in Southern Ocean Systems.
This working group was approved in 2015 and is proposed to finish in 2019.
The full terms of references for this working group can be found here.
Coordinate research and development into the use of satellite remote sensing with the ultimate aims of (i) undertaking a global census of Antarctic pack-ice seals and (ii) implementing regular monitoring of pack-ice seals in the Southern Ocean, which provide an integrated signal on the state of the Southern Ocean;
Hold annual meetings of the working group, and to source funding to enable this;
Develop standardised methods that are easily understood by different Antarctic stakeholders, including policy stakeholders, which are repeatable, and easily transferred to other research teams looking to contribute to these surveys;
Determine the optimal division of labor to achieve regular continental-scale surveys. One option might be to perform regional assessments that could be combined post-facto into a global assessment. The advantage of this approach is that it naturally accommodates regionally specific approaches that account for differences in satellite coverage, regional climate differences e.g. cloud cover, sea ice conditions, the spatial distribution and composition of the seal assemblages, and the capacity to perform ground truth surveys;
Develop analytical/statistical procedures for estimating seal abundance and associated estimates of error, with particular consideration of estimator bias and precision;
Establish how population estimates and other products would be delivered to end users such as CCAMLR, SOOS, and SCAR.
Planned Products and Outcomes
- Develop automated algorithms for reliably detecting seals in satellite images with ground truthing lgorithms against known presence and absence of seals and ultimately measures of estimate uncertainty. In addition, develop and test the applicability of automated and semi-automated routines using different resolution of VHR imagery.
- Design and implement a robust survey design including (a) the number of images needed to sample the region and to minimise uncertainty of the resulting estimates and (b) the spatial sampling arrangement, taking into account the stratified nature of the distribution of pack-ice seals with respect to the continental shelf.
- If outcomes 1 and 2 are successfully achieved; assess the state of the krill predator (pack-ice seals) component of the Antarctic marine ecosystem.
- Develop a habitat model for pack-ice seals based on seal distribution and physical parameters including sea-ice concentration (taken directly from the images), bathymetry and a suite of derived variables such as age of ice.
Contributing Modelling and Field Activities