CIMMSE TC Wind Group Conference Call Wednesday, July 10 th , 2013 11:00 AM
Dec 30, 2015
CIMMSE TC Wind GroupConference Call
Wednesday, July 10th, 201311:00 AM
Agenda
• Modified Rankine error function (MREF) survey results
• Modeling Update– First WRF-LES simulation problems– Potential solutions suggested by collaborators
MREF Survey Results
• Collaborators were asked to submit comments based on the images posted at:https://sites.google.com/a/ncsu.edu/modified-_rankine_improvements/
• Two NWS WFOs provided comments based on the supplied images (at least four forecasters within the two WFOs)
MREF Survey Results • Consensus that the MREF interpolated winds were better than
the currently used MR winds
– MREF: Strong reduction in wind speeds in the left storm quadrants consistent with forecaster experiences for storms paralleling the coastline
– MREF: Improvements in right front quadrant (strong reduction in wind speeds just outside the eye wall; larger radius of moderate wind field near storm periphery are consistent with forecaster experiences)
– In general, MREF decreases wind speeds for much of storm—consistent with NDFD verification results as a necessary measure
– “Improvements in WFO-to-WFO consistency while also retaining the intent of NHC forecasters”
MREF Survey Results • There were a few storms that MR performed better than
MREF
– Sometimes MR does better in the 40-60 km distance from storm center area (Felix, Paloma, Rina)
– Typically storms with “atypical” storm/wind structure have MREF failing
• Examples: Felix (2007), Earl (2010)
• Suggestion: incorporate into TCMWindTool and then allow WFOs in region to collaborate and decide which interpolation method to use for the given storm)
Modeling Setup
• 5 Domains, with innermost two domains having no PBL scheme
• Use of 1/16° daily SST analyses, 0.5° six hourly GFS analyses for boundary conditions
• Grid Spacing (consistent with Zhu et al. 2008):– D1: 8.1 km D3: 900 m D5: 100 m– D2: 2.7 km D4: 300 m
10-m wind speeds (knots) for domain 2
Initial WRF-LES Simulation Problem• After a few hours, an anomalous, unphysical
wind maximum develops within the WRF-LES domain 4 – Potential causes (after discussing with Dr. Sukanta
Basu)• Two-way nesting used—NCAR does not recommend for
WRF-LES domains—use one-way nesting for WRF-LES domains and two-way for WRF-mesoscale domains• WRF-LES domain was situated over the land-sea
interface (suggested to use WRF-LES domain only over ocean or land)
Modeling Goals/Focuses• Use Irene (2011) as a first case study
– Conducted 24 hour WRF-LES simulation near location of landfall beginning at 0000 UTC 27 August
– Examine large eddy circulations within the boundary layer– Conduct TKE budget to explore relative magnitude of
contributions of the various terms– Examine gust factors at various locations within inner domain
(compare to the climatology analysis we have conducted)• Sensitivity studies
– Lower level stability– Storm intensity
Discussion
Next Conference Call: Wednesday, August 14th, 2013 11 AM