[Eoas-seminar] MET Special Seminar Friday March 6th at 2:00

eoas-seminar at lists.fsu.edu eoas-seminar at lists.fsu.edu
Thu Feb 26 12:09:49 EST 2015


Meteorology Special Seminar

Dr. Mike Fiorino



NOAA ESRL

Boulder, CO


Deterministic Tropical Cyclone Genesis in a Suite of Operational Global NWP Model

Friday, March 6th
2:00 PM



Werner A. Baum Seminar Room (353 Love Building)
(Please join us for refreshments served outside room 353 Love @ 1:30 PM)


ABSTRACT
The improvement in tropical cyclone (TC) track forecasts by global models over the last 25 years has recently (circa 2010) reached the point that the models are now accurately forecasting TC genesis.  In contrast to probabilistic TC development products, this study focuses on deterministic, yes-or-no forecasting by the high-resolution global models.

Their are two requirements for any study of TC genesis: 1) a data set of tropical disturbances that develop (or not) into TCs; and 2) a definition of genesis.  We use operations to satisfy these two requirements.

The TC warning process at both JTWC and NHC almost always begins by starting an INVEST (9X ATCF ID).  An INVEST is a system/disturbance of interest that forecasters suspect may develop into a TC.  These systems are called pre/potential TCs or pTCs and the initiation of a pTC starts satellite reconnaissance and the generation of TC-specific products and imagery.  However, pTC data are not well maintained by JTWC/NHC but through data archaeology and local maintenance, a 15-y global data set has been constructed and curated.  The demographics of pTC/TCs in the Western north Pacific (WPAC) will discussed, but the main finding is that ~ 15% of the 100+ pTC develop in a WPAC season.  Moreover, the mean time between the initiation of a pTC and formation into TC is about 60 h.

Both JTWC and NHC are required to issue warnings (called advisories at NHC) for all TCs regardless of intensity. We define a 30-h genesis window centered on the first warning.  Comparison of the time of first warning to the time of the first TC position in the post-season reanalyzed best track shows the operational definition is at most 6-h difference.  The 30-h window allows a model that is run only at 00/12 UTC to have three chances to forecast a TC in the genesis window.  A correct forecast ('yes') is defined as at least one run out three with a 'significant' model TC near the genesis position.  A signifant model TC is measured with a duration-intensity unit called the 'scaled tropical depression days' or sTDd and is the sum of model intensity (maximum surface wind speed from the tracker) divided by 25 kt times the time period (6 h).  Model TCs (from the tracker) that cannot be associated with a pTC or TC are spurious and are thus called 'spuricanes'.  The formation and distribution of spuricanes is related to model physics, in particular the ratio of convective to total rainfall over the tropical oceans.  In 2013, all models in the suite that includes: 1) ECMWF; 2) UKMO; 3) NCEP/ESRL; 4) CMC and 5) NAVGEM correctly predicted genesis in WPAC at day two and ~ 80% at day five.  Results for 2014 in the Pacific
and Atlantic basins will be given to see if there are trends in genesis forecast skill.Both JTWC and NHC are required to issue warnings (called advisories at NHC) for all TCs regardless of intensity. We define a 30-h genesis window centered on the first warning.  Comparison of the time of first warning to the time of the first TC position in the post-season reanalyzed best track shows the operational definition is at most 6-h difference.  The 30-h window allows a model that is run only at 00/12 UTC to have three chances to forecast a TC in the genesis window.  A correct forecast ('yes') is defined as at least one run out three with a 'significant' model TC near the genesis position.  A signifant model TC is measured with a duration-intensity unit called the 'scaled tropical
depression days' or sTDd and is the sum of model intensity (maximum surface wind speed from the tracker) divided by 25 kt times the time period (6 h).  Model TCs (from the tracker) that cannot be associated with a pTC or TC are spurious and are thus called 'spuricanes'.  The formation and distribution of spuricanes is related to model physics, in particular the ratio of convective to total rainfall over the tropical oceans.  In 2013, all models in the suite that includes: 1) ECMWF; 2) UKMO; 3) NCEP/ESRL; 4) CMC and 5) NAVGEM correctly predicted genesis in WPAC at day two and ~ 80% at day five.  Results for 2014 in the Pacific and Atlantic basins will be given to see if there are trends in genesis forecast skill.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fsu.edu/pipermail/eoas-seminar/attachments/20150226/2d95d722/attachment-0001.html>


More information about the Eoas-seminar mailing list