Tropical Cyclones

From the oil rich Gulf of Mexico or the South China Sea, to the

population centers along the Atlantic coast or Bay of Bengal,

CFAN’s forecasts can help you anticipate impacts further in advance

and with greater accuracy and precision. 


TropiCast empowers our clients in government and industry to visualize and better understand the overall risk of impact from a hurricane or tropical cyclone in the Atlantic, North Indian Oceans, and West Pacific. TropiCast can support your decision making for: offshore resource operation and safety, utilities operations and recovery, energy demand and trading, insurance and reinsurance needs, emergency management, ocean routing and fleet management, travel and tourism impacts, and supply chain interruption and restoration. 


Building upon our innovative tropical cyclone research [link], CFAN has been defining the state-of-the-art in tropical cyclone forecasting since 2007.  

Our extended-range tropical cyclone forecast products out to 15 days are unmatched in the weather services industry. TropiCast’s unique forecasts are produced using CFAN’s proprietary algorithms applied to global model output ensembles in near real time (referred to as ‘calibration’), which provide more accurate forecasts than raw (uncalibrated) model output with features including:

  • calibrated tracks from global model ensembles, which provide up to 3 days of additional lead time relative to raw model output tracks

  • probability forecasts of tropical cyclone formation (genesis) that have skill 3-7 days in advance

  • dynamic cone of uncertainty for tracks forecasts which provides more realistic track uncertainty at extended forecast horizons than raw model output.

  • probability forecasts of hurricane intensity that provide a more realistic assessment of intensity uncertainty


For each active system – a named storm or National Hurricane Center 'invest' – the active system page includes:

  • An interactive feature that allows you to overlay track and intensity forecasts from the global models plus NHC’s statistical and regional dynamical models

  • Realtime track and intensity verification for the global and regional dynamical models, for the past 4 days

  • Track probabilities created from synthetic tracks using a proprietary Monte Carlo analysis of the the ECMWF and NOAA global models, providing a dynamic cone of track uncertainty

  • Probability forecasts of maximum and intensity and minimum pressure

  • Forecasts of Integrated Kinetic Energy (IKE) and Cyclone Damage Potential (CDP) to help anticipate relative landfall damage

  • Rapid Intensification forecasts -  NHC forecasts plus forecasts from CFAN's proprietary Rapid Intensification Index

  • Visible and near infrared satellite imagery, with overlay from the Global Lightning Mapper


The images below illustrate the active system forecast for Hurricane Dorian (8/30/19 12Z) – at a critical point in the forecast when catastrophic impacts were anticipated for the northern Bahamas and the forecast models showed substantial disagreement on a Florida landfall. ECMWF and NCEP model data depicted in the images have been improved by CFAN’s calibration.


Atlantic tropical cyclones


CFAN has been providing tropical cyclone forecasts in the North Atlantic since 2007.  CFAN’s probabilistic forecasts of Atlantic tropical cyclone tracks and intensity consistently outperform both government and market forecasts beyond a 2 day time horizon, showing track skill within 300 miles out to 10 days (even before the tropical cyclones actually form).  

CFAN’s superior track forecast is illustrated by comparing CFAN’s forecast for Hurricane Sandy with raw tracks from ECMWF and GEFS, 7 days before landfall.

CFAN Forecast

ECMWF Forecast

GEFS Forecast

This view of the TropiCast dashboard shows forecasts for Hurricane Irma (2017) 6 hours prior to formation and 12 days before landfall on the southwest coast of FL. 

2017 track verification analysis.  At five days lead time, the average track error for the 2017 NHC forecasts was 170 miles.  CFAN’s five day track forecast error was smaller than the official NHC forecast by an average of 39 miles (26%). At 10 days lead-time  the average track error for CFAN’s forecast was 287 miles. CFAN's tracking and calibration improves upon the official ECMWF track forecasts by 7 miles at 5 days and 69 miles at 10 days.  Download the full Report on Track Verification Analysis for the 2017 Atlantic Hurricane Season  [link]

Detailed information on the extent and severity of winds and rainfall in an approaching hurricane assists property owners, facility managers and emergency personnel in preparations and planning aimed at minimizing losses from an eventual impact.  


CFAN provides customized high-resolution (5 km) ensemble wind forecasts to support damage estimates and emergency management. CFAN’s predictions of landfall impacts of hurricane winds use an adaptive statistical-dynamical technique to statistically downscale the coarse resolution surface winds from the global model ensembles plus the National Hurricane Center models.  The forecasted wind speeds from the global models are calibrated using CFAN’s calibrated intensity forecasts and initial observations of the radius of hurricane winds.

The movie below portrays CFAN’s prediction of high-resolution landfall winds for Hurricane Hermine (2016), 6 days prior to landfall and before it became a named storm.


Landfall impacts

All Videos

All Videos

Watch Now

Verification analysis of 2016/2017 FL landfall winds.  CFAN's high-resolution landfall wind forecasts out to 72 hours are evaluated against the operational analyses that are downscaled to the same resolution as the forecasts [link]

CFAN provides an interactive tool that allows the user to dynamically zoom into the region of interest to display the forecasts: 

  • Tracks – overlay CFAN’s calibrated ECMWF and GFS tracks, NHC tracks, other global models, to assess specific landfall scenarios

  • Winds -  calibrated and downscaled winds for each ECMWF and GFS ensemble member; a radial wind model based on forecast parameters from the National Hurricane Center

  • Wind gusts – CFAN’s calibrated ECMWF wind gust forecasts

  • Rainfall – 6 hr, 24hr and 72 hr accumulations, 5 mile resolution.  Based on ECMWF forecasts.


CFAN’s landfall impact tool is illustrated for forecasts of Hurricane Dorian (2019):


These forecasts are also provided as a data stream to interface with models of damages to client-specific assets and operations.

Seasonal forecasts

Seasonal forecasts of hurricane activity benefit financial decision making and basic risk awareness from tropical storms.


CFAN’s seasonal forecasts for Atlantic tropical cyclone activity are based on a breakthrough in understanding of the impact of global climate dynamics on Atlantic hurricane activity. CFAN’s forecasts provide greater prediction skill for years particularly when there is no clear signal from ENSO (El Nino/La Nina).


CFAN has identified skillful new predictors for the annual basin-scale Accumulated Cyclone Energy (ACE), which integrates the duration and intensity of Named Storms.number of U.S. landfalling hurricanes and the seasonal ACE. North Atlantic hurricane activity varies substantially from year to year, including 5-fold variations in ACE. 



CFAN issues seasonal hurricane forecasts for the Atlantic in December, April, June and August.  Seasonal forecasts for the West Pacific are issued in February and July.  Forecast reports are available only via paid subscription.  Contact us

In addition to the North Atlantic, CFAN provides tropical cyclone forecast and extended range outlooks for the following regions:


  • West Pacific (Northern hemisphere)

  • East Pacific (Northern hemisphere)

  • North Indian Ocean



Global Tropical Cyclones

Trial subscription


CFAN offers trial subscriptions to TropiCast during the months of May, June, October and November to qualified individuals and organizations. 

To inquire about a trial subscription [link]