|
Tropical Cyclone Research Team
Chris Velden
Tim Olander
Dave Stettner
Derrick Herndon
Tony Wimmers
Sarah Griffin Jeff Hawkins
Remote Collaborators
Jason Dunion (NOAA/HRD)
|
|
The main questions facing tropical cyclone (TC) forecasters are accurately determining
the current intensity of the storm, the future intensity of the storm, and where the
storm will be in the future. Data obtained from satellites are ideal for helping to provide
answers to these questions due to their ability to provide nearly constant and total coverage
of the tropics in space and time.
The CIMSS Tropical Cyclone webpage helps achieve these goals by providing near real-time imagery,
derived atmospheric analysis products, and TC intensity estimates from a variety of
different satellite platforms for global analysis of TCs and their surrounding environments.
Many of the products from CIMSS are developed specifically for use by TC forecasters worldwide
to provide unique information in support of their specific TC forecasting missions.
The CIMSS Tropical Cyclone group does not produce forecasts or warnings for any TCs, instead
we provide unique information to the organizations and forecasters who do.
|
|
|
Scroll over tabs below for brief description of each project, the primary investigators, and a link
to project homepages
The Advanced Dvorak Technique (ADT) provides real-time TC intensity estimates utilizing
geostationary, infrared satellite imagery. The algorithm based on the Subjective Dvorak
Technique, utilized at all TC forecasting centers worldwide, but it eliminates the inherent
subjectivity of the technique. ADT intensity estimates can be produced
in a completely automated manner over the entire TC lifecycle.
Primary Investigators :
Tim Olander,
Chris Velden
Project Homepage
Atmospheric motion vectors (AMV) are derived from a sequence of satellite imagery. Various channels
can be used to calculate wind vectors at different atmospheric levels. The wind vectors can be used qualitatively by
TC forecasters help visualize the wind field around TCs, as well as quantitatively within numerical weather prediction
models. Atmospheric analysis products, such as wind shear, vorticity, and convergence, can be also be derived from
these wind fields.
Primary Investigators :
Chris Velden,
Steve Wanzong,
Dave Santek,
Dave Stettner
Project Homepage
The Morphed Integrated Microwave Imagery at CIMSS (MIMIC) product provides
a radar-like visualization of the evolution of TC structure using advanced image morphing
algorithms. The final product is a synthetically-derived animation at 15 minute intervals
between two authentic passive microwave satellite images. The morphing routines utilize
an advection/rotation scheme based on the current reported TC wind speed estimates.
Primary Investigators :
Tony Wimmers,
Chris Velden
Project Homepage
The Morphed Integrated Microwave Imagery at CIMSS (MIMIC) product
presents total precipitable (TPW) water over the ocean, retrieved from SSMI and AMSR-E
microwave sensors. The final product is an hourly composite of many swaths of TPW retrievals,
advected to the required time using 1000-600 hPa winds from the GFS model.
Primary Investigators :
Tony Wimmers,
Chris Velden
Project Homepage
The CIMSS AMSU algorithm utilizes observed microwave radiation
from several different wavelengths to measure storm-related atmospheric warming during
TC events at different heights within the storm. The amount of anamolous warming
within the TC core is determined, after adjusting for the AMSU scanning geometry
and resolution effects, and is then related to a storm intensity.
Primary Investigators :
Derrick Herndon,
Chris Velden
Project Homepage
The CIMSS Satellite Consensus (SatCon) algorithm blends individual
TC intensity estimates derived from the ADT, CIMSS AMSU, and CIRA AMSU routines to produce
ensemble TC intensity estimate. The individual intensity estimates are combined using a
weighting scheme which effectively maximizes/minimizes the influence of each estimate
based upon the strengths/weaknesses of each technique for the currently observed TC
environment and structure.
Primary Investigators :
Chris Velden,
Derrick Herndon,
Tim Olander
Project Homepage
The imagery and derived analysis products displayed allow
users to identify and track the atmospheric feature called the Saharan Air Layer (SAL).
This feature consists of a very dry and dusty low-level layer of air formed over the Saharan Desert
of Africa which can spread over large areas the tropical Atlantic Ocean. The air within
the SAL can drasitcally affect TC intensity and inhibit formation and/or development of
TCs over the entire Atlantic Ocean basin.
Primary Investigators :
Jason Dunion,
Chris Velden
Project Homepage
This project employs an objective satellite-based tropical overshooting
top (TOT) detection algorithm in the tropical Atlantic and Eastern Pacific Oceans. The TOTs
are used as a proxy indicator to analyze vigorous convection trends in tropical disturbances
that may become incipient tropical cyclones. Preliminary research has indicated some skill in
using the TOT product as a predictor of TC genesis and rapid intensification in the Atlantic Ocean.
Primary Investigators :
Sarah Griffin,
Chris Velden
Project Homepage
Rapid-Scan Atmospheric Motion Vectors provide a unique opportunity to observe
high spatial and temporal resolution wind fields in the upper-levels
of tropical cyclones. The upper-level analyses are crucial for understanding
the storm's outflow and mass flux. Thus, studies are underway to investigate
the connection and influence of these upper-level winds to tropical cyclone
intensity change.
Primary Investigators :
Chris Velden,
John Sears
Project Homepage
The ARCHER product and supporting real-time page are designed to aid tropical cyclone forecasters in sifting through the increasing wealth of relevant satellite data to quickly and objectively arrive at key TC characteristics. It is not intended to substitute for the skill of experienced analysts, but rather to increase the efficient use of analysis time.
Primary Investigators :
Tony Wimmers,
Chris Velden
Project Homepage
The AI-enhanced Advanced Dvorak Technique (AiDT) is a machine learning (ML) extension of the ADT which provides tropical cyclone intensity estimates around the globe. The AiDT uses ADT history file outputs to feed into the ML model to derive a wind speed estimate using a simple one-layer model configuration.
Primary Investigators :
Tim Olander,
Tony Wimmers,
Chris Velden
Project Homepage
The AI Rapid Intensification Index (AI-RI) predicts TC rapid intensification using geostationary infrared imagery and scalar environmental predictors in a convolutional neural network. AI-RI predicts the probability of RI over 12, 24, 36, 48, and 72 hour lead times, similar to the operational SHIPS prediction. This is available in the North Atlantic and East Pacific basins.
Primary Investigators :
Sarah Griffin,
Tony Wimmers,
Chris Velden
Project Homepage
D-MINT (Deep Multispectral INtensity of TCs estimator, formerly DeepMultiNet) provides real-time TC intensity estimates using microwave (SSMIS, AMSR-2, and GMI) and geostationary infrared satellite imagery together in a convolutional neural network. This is available worldwide.
Primary Investigators :
Tony Wimmers,
Sarah Griffin,
Chris Velden
Project Homepage
Coming Soon
Primary Investigators :
Jason Dunion,
Chris Velden
Project Homepage
The D-PRINT (DeeP IR INtensity of TCs, formerly OPEN-AIIR) makes real-time TC intensity estimates using geostationary infrared satellite imagery in a convolutional neural network. It currently predicts TC intensity every hour and uses satellite imagery over the previous 9 hours in its estimation. This is available worldwide.
Primary Investigators :
Tony Wimmers,
Sarah Griffin,
Chris Velden
Project Homepage
|
|
