Improving light‐based geolocation by including sea surface temperature

TLDR

The study builds on a Kalman filter–based state‑space model for light‑level geolocation. It aims to incorporate sea surface temperature measurements into PSAT light‑level geolocation estimates. The method integrates SST into the Kalman filter alongside light‑derived longitude and latitude, is validated on PSAT‑tagged drifter buoys, and applied to blue shark tracks. Adding SST yields more probable tracks with reduced prediction variance compared to light‑only estimates.

Abstract

Abstract An approach to integrate sea surface temperature (SST) measurements into estimates of geolocations calculated by changes in ambient light level from data downloaded from pop‐up satellite archival tags (PSAT) is presented. The model is an extension of an approach based on Kalman filter estimation in a state‐space model. The approach uses longitude and latitude estimated from light, and SST. The extra information on SST is included in a consistent manner within the milieu of the Kalman filter. The technique was evaluated by attaching PSATs directly on thermistor‐equipped global positioning system drifter buoys. SSTs measured in the PSATs and drifter buoy were statistically compared with SSTs determined from satellites. The method is applied to two tracks derived from PSAT‐tagged blue sharks ( Prionace glauca ) in the central Pacific Ocean. The inclusion of SST in the model produced substantially more probable tracks with lower prediction variance than those estimated from light‐level data alone.