APIXQUI is a Mexican interinstitutional initiative aimed at strengthening ocean observations and forecasting capabilities for extreme ocean-atmosphere events. The initiative emerged in response to the devastating impacts of Hurricane Otis (2023), which exposed important limitations in observing and forecasting rapidly intensifying tropical cyclones. The first phase of the project focuses on the poorly sampled tropical Mexican Pacific, where CICESE contributes through autonomous underwater gliders and metocean buoys, as well as coupled ocean-atmosphere-wave modelling, numerical forecasting and emerging artificial intelligence-based prediction methodologies. The observations will support tropical cyclone forecasting while also helping to understand how upper-ocean heat content, stratification and air-sea interactions modulate tropical cyclone rapid intensification and tracks. At the same time, they will contribute to investigating the effects of other phenomena, including El Niño events and marine heatwaves. The observations and forecasting products generated through the project will be openly available to the scientific and operational communities, supporting research and applications beyond tropical cyclone forecasting.
[Fig 1 (right): Ocean heat content (OHC) on 12 August 2026 produced by the Glider Oceanographic Monitoring Group (GMOG, https://gliders.cicese.mx/) from Copernicus Marine Service absolute dynamic topography (ADT) maps. The OHC–ADT relationship was obtained from a statistical regression using historical Argo observations and Copernicus Marine Service ADT data. The thick contour indicates the 40 kJ cm⁻² OHC threshold. The panels show OHC along the trajectories of gliders SG625 and SG623, comparing estimates derived from in situ temperature observations with those obtained from the OHC maps. Shaded areas indicate the root-mean-square error of the daily estimates | Credit: Glider Oceanographic Monitoring Group (GMOG), CICESE]
Every year, tropical cyclones develop over both the eastern North Pacific and the North Atlantic basins. With more than 11,000 km of coastline along the Pacific Ocean, the Gulf of Mexico and the Caribbean Sea, Mexico is highly exposed to these systems. On average, nearly 30 named tropical cyclones form annually across these two basins, and several make landfall or pass close enough to affect human life, coastal communities, marine operations and critical infrastructure.
[Figs. 2a & 2b. Autonomous underwater glider deployed during an APIXQUI ocean observing mission in the tropical Mexican Pacific. The glider repeatedly measures temperature and salinity throughout the water column. Credit: Glider Oceanographic Monitoring Group (GMOG), CICESE.]
Despite the significant advances achieved in numerical weather prediction over recent decades, important gaps remain in our scientific understanding of the role of the ocean in tropical cyclone development, rapid intensification, and ocean-atmosphere interactions, as well as in our ability to translate that understanding into improved forecasts. Hurricane Otis (2023), which intensified from a tropical storm to a Category 5 hurricane in less than 24 hours before making landfall near Acapulco, clearly exposed the limitations that still exist in forecasting rapid intensification events. The hurricane caused major loss of life, extensive damage to infrastructure, and severe economic losses. Recent studies have increasingly demonstrated the value of ocean observations for improving forecast skill and quantifying the influence of the ocean on tropical cyclone evolution. In response to these scientific and operational challenges, the Government of Mexico, through the Ministry of Science, Humanities, Technology and Innovation (SECIHTI), launched APIXQUI (the Nahuatl word for water sentinel), an interinstitutional effort to strengthen national ocean observations and forecasting capabilities for extreme ocean-atmosphere events affecting the Mexican coasts.
Although the initial motivation for APIXQUI was to improve tropical cyclone forecasting, the project addresses a broader range of scientific questions. Improving forecasts requires improving observations, but these observations also provide the opportunity to investigate physical processes that are still not fully understood. Processes such as the role of upper-ocean heat content, vertical stratification and air-sea interactions in tropical cyclone evolution remain important research questions. The observations will also contribute to investigating other oceanic and hydrometeorological phenomena affecting Mexican waters, including El Niño–Southern Oscillation variability, particularly El Niño events, marine heatwaves and other extreme events.
[Figs. 3a and 3b: Metocean buoy to be deployed as part of the APIXQUI observing network, measuring atmospheric and oceanographic conditions at the air-sea interface | Credit: CANEK Group, CICESE.]
APIXQUI is coordinated by SECIHTI and integrates the expertise and operational capabilities of the National Autonomous University of Mexico (UNAM), CICESE, the Mexican Navy (SEMAR), the National Meteorological Service (SMN) and other national partners. The initiative builds upon existing national capabilities by strengthening observational infrastructure, expanding forecasting capabilities and promoting closer collaboration between research institutions and operational agencies.
The first phase of the project focuses on the tropical Mexican Pacific, one of the most active regions for tropical cyclone genesis and rapid intensification, yet one that remains poorly sampled by in situ ocean observations. Expanding observations in this region addresses one of the largest observational gaps surrounding the Mexican coasts. Within this framework, CICESE contributes to the ocean observing component through the deployment of complementary observing platforms designed to monitor the upper ocean in near-real time. Autonomous underwater gliders repeatedly sample temperature and salinity throughout the water column in regions of tropical cyclone genesis and rapid intensification, while a new network of metocean buoys continuously measures atmospheric and oceanographic variables at strategically selected locations, covering key regions along tropical cyclone tracks.
These observations are complemented by coupled ocean-atmosphere-wave modelling, numerical forecasting, and new forecasting methodologies based on artificial intelligence, supported by high-performance computing capabilities at CICESE. Together, these components allow the observations to be used not only for forecasting, but also to investigate how upper-ocean heat content, stratification and air-sea interactions influence tropical cyclone rapid intensification and tracks. The same observations can also be used to evaluate and develop forecasting systems and to investigate ocean and atmospheric variability beyond tropical cyclones.
An important component of APIXQUI is that the observations and forecasting products generated through the project will be made openly available to the scientific and operational communities. In this way, strengthening Mexico’s national observing and forecasting capabilities can go hand in hand with making these observations available as a resource for the wider scientific and operational community.
[Fig. 4: Hurricane Otis as a Category 5 hurricane approaching the coast of Guerrero, Mexico, on 25 October 2023 at 04:30 UTC. Infrared imagery from the GOES-East satellite (Band 13, 10.3 µm) shows the highly organized structure and intense deep convection surrounding the eye. Credit: NOAA/NESDIS/STAR, GOES-East.]
Figure 1. Ocean heat content (OHC) on 12 August 2026 produced by the Glider Oceanographic Monitoring Group (GMOG, https://gliders.cicese.mx/) from Copernicus Marine Service absolute dynamic topography (ADT) maps. The OHC–ADT relationship was obtained from a statistical regression using historical Argo observations and Copernicus Marine Service ADT data. The thick contour indicates the 40 kJ cm⁻² OHC threshold. The panels show OHC along the trajectories of gliders SG625 and SG623, comparing estimates derived from in situ temperature observations with those obtained from the OHC maps. Shaded areas indicate the root-mean-square error of the daily estimates.