T opical Cyclone Rada A chi e o Dopple Analyses wi h Recen e ing (TC-RADAR)
Michael Fische 1, Paul Reaso 2, Robe Roge s2, and John Gamache2
1Rosens iel School o Ma ine, A mosphe ic, and Ea h Science, Uni e si y o Miami, Miami, FL
2NOAA/OAR/A lan ic Oceanog aphic and Me eo ological Labo a o y, Miami, FL
C ea ed: 3 Ap il 2020
Las upda ed: 26 Augus 2025
Cu en da abase e sion: 3m
1. In oduc ion and mo i a ion:
NOAA’s Hu icane Resea ch Di ision (HRD) ou inely collec s ai bo ne ada obse a ions o opical
cyclones (TCs) in he No h A lan ic, and occasionally, he eas e n and cen al No h Paci ic basins.
Onboa d NOAA’s P3 ai c a a e h ee ada s, loca ed on he nose, lowe uselage, and ail o he ai c a
( o mo e in o ma ion, see: h ps://www.aoml.noaa.go /h d/abou _h d/HRD-P3_ ada .h ml). The ail
Dopple ada (TDR) is an X-band ada ha scans in bo h he o e and a di ec ions, allowing
obse a ions o he h ee-dimensional kinema ic s uc u e o a TC (Gamache e al. 1995). Typically, a
gi en mission will ha e 3–4 passes h ough he cen e o he s o m. Fo each cen e pass h ough he
s o m, a TDR analysis is c ea ed. In ypical ope a ing condi ions, ada obse a ions only ex end ou wa d
~50 km om he ai c a , limi ing he azimu hal co e age o he s o m. An example o he co e age he
TDR p o ides o a se o passes is shown in Fig. 1 below.
Fig. 1. a) TDR analyses o e lec i i y (shaded; dBZ) and s o m- ela i e wind (ba bs; k ) a a heigh o 7 km
o he 100828I1 mission in o T opical S o m Ea l. The 2-km and 7-km TC cen e s a e deno ed by he cyan
ci cle and squa e, espec i ely. Analysis is alid a 2132 UTC 28 Augus 2010. Coinciden sa elli e-de i ed
in a ed b igh ness empe a u es (°C) a e shown in he g ayscale shading. The en i onmen al e ical
wind shea di ec ion is shown by he ed a ow in he op-le inse . b–d) As in a), bu o analyses alid a
2251 UTC 28 Augus , 0007 UTC 29 Augus , and 0131 UTC 29 Augus , espec i ely.
As can be seen om Fig. 1, he azimu hal co e age o obse a ions om hese “swa hs” o da a is limi ed
o egions nea he ligh ack. To emedy his issue, he ai c a usually conduc s mul iple passes
h ough he cen e o he s o m a a ying azimu hal angles. These swa hs o da a can hen be
combined, usually by a e aging all swa h analyses c ea ed om a single ligh in o a “me ged” analysis.
The g ea e azimu hal co e age o obse a ions p o ided by me ged analyses ha e been shown by
p e ious s udies o be use ul o examina ions o symme ic and asymme ic o ex cha ac e is ics. An
example compa ison o TDR swa h analyses o me ged analyses is p o ided in Fig. 2, o a ligh in o
Hu icane Do ian (2019).
Fig. 2. a) S o m-cen e ed, mo ion- ela i e, TDR-de i ed ho izon al wind speed (shaded; m s-1) a a heigh
o 2 km ob ained om he 1324 UTC 1 Sep embe cen e pass in o Hu icane Do ian (2019) as pa o he
20190901H1 mission. The app oxima e ligh ack is shown by he black-dashed line, wi h he du a ion
o he obse ing pe iod shown in he legend. This pass se es as a ep esen a i e example o swa h da a.
b) As in a), bu o he co esponding me ged analysis using swa h da a om all h ee cen e passes in
he 20190901H1 mission. The co esponding ligh ack o each TDR swa h analysis is shown by a
unique dash s yle, as deno ed in he legend. Figu e adap ed om Fig. 1 o Fische e al. (2022).
To acili a e u u e esea ch o TC o ex and con ec i e cha ac e is ics, he TC-RADAR da abase p o ides
a pla o m whe e pos -p ocessed swa h and me ged analyses a e combined in o wo homogeneous da a
iles, wi h ele an me ada a pe aining o bo h he TC bes - ack in o ma ion and he TC en i onmen ,
as diagnosed by he ope a ional S a is ical Hu icane In ensi y P edic ion Scheme (SHIPS) model. Such a
o ma eadily allows o s o m- o-s o m compa isons o clima ological analyses. I is ecommended ha
swa h analyses a e used o examine he con ec i e cha ac e is ics o he s o m, as he a e aging p ocess
in he me ged analyses will smoo h ea u es ha a e sampled by mo e han one pass.
In o al, TC-RADAR con ains 1,510 swa hs o TDR obse a ions and 436 me ged analyses o s o ms
occu ing in he No h A lan ic, eas e n No h Paci ic, and cen al No h Paci ic basins. Obse a ions span
all poin s o he TC li ecycle, anging om p e-genesis dis u bances, o ma u e hu icanes, o s o ms
nea ing ex a opical ansi ion.
2. Recen e ing p ocess
One o he bene i s o using TC-RADAR is ha analyses a e eadily a ailable on s o m-cen e ed g ids. The
TC cen e is iden i ied using a weigh ed ci cula ion maximiza ion (WCM) me hod (Fische e al. 2024). To
summa ize, he TC cen e is iden i ied as he analysis g id poin ha bes ma ches a pe ec ly symme ic
o ex o pu ely cyclonic low. The quali y o he ma ch is de e mined ia a cos unc ion ha conside s
he adial dis ance o analysis g id poin s om he TC cen e and he wind speed.
In a bi mo e de ail, in he cen e - inding p ocess, he s o m- ela i e wind di ec ion a each g id poin is
compa ed o a low ha is pu ely angen ial abou a hypo he ical TC cen e . This p ocess is epea ed o a
se ies o hypo he ical TC cen e loca ions. The bes ma ch is iden i ied as he loca ion ha yields he
lowes mean di e ence be ween he obse ed wind di ec ion and he pu ely angen ial low. The “e o ”
be ween he obse ed wind di ec ion and he pu ely angen ial low a e weigh ed by bo h dis ance om
he TC cen e and he wind speed, whe e as e winds and poin s close owa d he TC cen e a e
weigh ed mo e s ongly han g id poin s wi h ela i ely weak wind speeds and/o poin s a he om he
TC cen e . This p ocess is pe o med o all swa h and me ged analyses. In he cons uc ion o me ged
analyses, he TC cen e is de e mined a a heigh o 2 km o each swa h analysis. Then, each swa h
analysis o a gi en ligh is a e aged abou he espec i e 2-km TC cen e es ima e o c ea e he me ged
analysis.
3. Da a o ma ing:
The e a e wo main ypes o da a iles. Bo h a e composed o ne CDF (.nc) iles. The i s ype con ains all
TDR swa h analyses, while he second ype con ains all TDR me ged analyses. Bo h da a ypes use
Ca esian g ids wi h a ho izon al g id spacing o 2.0 km and e ical g id spacing o 0.5 km. The e ical
ex en o he domain spans 0–18 km in heigh . P esen ly, he e a e wo swa h analysis and wo me ged
analysis TC-RADAR .nc iles, g ouped by consecu i e yea s (e.g., 1997–2019 and 2020–2024).
Swa h da a (e.g., 3m_1997_2019_xy_ el_swa h_ships.nc):
Swa h da a is ideal o examining ei he p ocesses e ol ing on sho ime scales (looking a pass- o-pass
changes) o examining single snapsho s in ime, such as he con ec i e cha ac e is ics o he s o m. The
ollowing obse ed a iables a e a ailable o he swa h da a:
● Zonal wind (m s-1)
● Me idional wind (m s-1)
● Ve ical eloci y (m s-1)
● Re lec i i y (dBZ)
● Wind speed (m s-1)
● Tangen ial eloci y (m s-1)
● Radial eloci y (m s-1)
● Ea h- ela i e zonal wind (m s-1)
● Ea h- ela i e me idional wind (m s-1)
● Rela i e o ici y (s-1)
● Di e gence (s-1)
Unless explici ly s a ed as “ea h- ela i e”, all ields a e s o m mo ion- ela i e. These a iables a e
a ailable on g ids ela i e h ee di e en ypes o TC cen e es ima es:
● Rela i e o a TC cen e based on he eal- ime cen e ix (e.g., “swa h_zonal_wind”)
● Rela i e o a TC cen e based on he ecen e ed TC posi ion (see sec ion 2), using a e e ence
heigh o 2 km (e.g., “ ecen e ed_zonal_wind”)
● Rela i e o a TC cen e based on he ecen e ed TC posi ion (see sec ion 2), a e e y e ical le el
(e.g., “ o al_ ecen e ed_zonal_wind”). In essence, his is a o ex il - ela i e amewo k.
These da a a e s o ed in ou -dimensional a ays o he o de : s o m index, la i ude/me idional
displacemen , longi ude/zonal displacemen , and e ical le el.
Swa h analyses also ha e ce ain o ex pa ame e s s o ed. These include:
● Ve ical p o iles o he adius o maximum wind
● Ve ical p o iles o o ex il magni ude, including ec o componen s
Me ged analyses (e.g., 3m_1997_2019_xy_ el_me ged_ships.nc):
He e, he me ged analyses we e cons uc ed using he mean alues o all swa hs o a gi en mission. As
a esul , me ged analyses a e p e e ed o analyzing o ex-scale cha ac e is ics o he TC, as
con ec i e-scale ea u es will be smoo hed in he a e aging.
Re-cen e ed ields in he me ged analyses a e cons uc ed in a wo-s ep p ocess. Fi s , swa hs a e
ecen e ed, ollowing he me hods desc ibed abo e. Second, a e ised cen e is es ima ed using he
me ged angen ial wind ield c ea ed om he ecen e ed swa h da a. Theo e ically, he g ea e
azimu hal co e age o he me ged analysis should gi e a mo e accu a e cen e loca ion han cen e
es ima es ob ained om single swa hs.
The ollowing obse ed a iables a e a ailable o he me ged da a:
● Zonal wind (m s-1)
● Me idional wind (m s-1)
● Ve ical eloci y (m s-1)
● Re lec i i y (dBZ)
● Wind speed (m s-1)
● Tangen ial eloci y (m s-1)
● Radial eloci y (m s-1)
● Ea h- ela i e zonal wind (m s-1)
● Ea h- ela i e me idional wind (m s-1)
● Rela i e o ici y (s-1)
● Di e gence (s-1)
Unless explici ly s a ed as “ea h- ela i e”, all ields a e s o m mo ion- ela i e. These a iables a e
a ailable on g ids ela i e h ee di e en ypes o TC cen e es ima es:
● Rela i e o a TC cen e based on he eal- ime cen e ix (e.g., “me ged_zonal_wind”)
● Rela i e o a TC cen e based on he ecen e ed TC posi ion (see sec ion 2), using a e e ence
heigh o 2 km (e.g., “ ecen e ed_zonal_wind”)
● Rela i e o a TC cen e based on he ecen e ed TC posi ion (see sec ion 2), a e e y e ical le el
(e.g., “ o al_ ecen e ed_zonal_wind”). In essence, his is a o ex il - ela i e amewo k.
These da a a e s o ed in ou -dimensional a ays o he o de : s o m index, la i ude/me idional
displacemen , longi ude/zonal displacemen , and e ical le el.
Me ged analyses also ha e ce ain o ex pa ame e s s o ed. These include:
● Ve ical p o iles o he adius o maximum wind
● Ve ical p o iles o o ex il magni ude, including ec o componen s
● Ve ical p o iles o local shea , including ec o componen s (use wi h g ea cau ion)
The p o iles o local shea a e h ee-dimensional a ays o he o de : s o m index, a e aging adius, and
e ical le el. He e he a e aging adius (“local_ adii”) spans 25–100 km in 5-km inc emen s. In o de o
he local shea o be compu ed, a leas 50% o all da a poin s in each geog aphic quad an (i.e.,
no hwes , sou hwes , sou heas , no heas ) mus ha e alid da a. O he wise, an es ima e o he local
shea is no compu ed. Consequen ly, i is possible ha o ce ain local_ adii an es ima e o he local
shea exis s, while i does no o o he local_ adii. Since azimu hal co e age gene ally dec eases a
la ge adii, smalle alues o local_ adii mo e equen ly ha e es ima es o he local shea . Fo e e ence,
he mean quad an co e age and minimum quad an co e age a e s o ed in he a iables o he
co esponding names.
4. SHIPS me ada a:
Selec SHIPS me ada a a e included wi hin bo h he swa h and me ged da ase s. All a iables a e
ob ained om SHIPS de elopmen al da a. The ollowing a iables a e cu en ly a ailable:
● T opical cyclone iden i ica ion code (“ cid_ships”)
● Bes - ack TC cen e la i ude (“la _ships”; deg ees)
● Bes - ack TC cen e longi ude (“lon_ships”; deg ees)
● Bes - ack maximum sus ained 10-m wind speed (“ max_ships”; m s-1)
● Bes - ack minimum cen al p essu e (“p es_ships”; hPa)
● Zonal componen o s o m mo ion (“mo ion_x_ships”; m s-1)
● Me idional componen o s o m mo ion (“mo ion_y_ships”; m s-1)
● Deep-laye (850–200-hPa) e ical wind shea magni ude (“shdc_ships”; k )
● Deep-laye (850–200-hPa) e ical wind shea di ec ion (“sddc_ships”; deg ees)
● Deep-laye (850–200-hPa) gene alized e ical wind shea magni ude (“sh g_ships”; k )
● Deep-laye (850–200-hPa) gene alized e ical wind shea magni ude (“shgc_ships”; k )
● Deep-laye (850–200-hPa) e ical wind shea magni ude in 200–800-km ann. (“sh d_ships”; k )
● Deep-laye (850–200-hPa) e ical wind shea di ec ion in 200–800-km ann. (“sh d_ships”; deg.)
● Mid-laye (850–500-hPa) e ical wind shea magni ude (“sh s_ships”; k )
● Mid-laye (850–500-hPa) e ical wind shea di ec ion (“sh s_ships”; deg ees)
● Lowe - oposphe ic (850–700-hPa) en i onmen al ela i e humidi y (“ hlo_ships”; %)
● Mid- oposphe ic (700–500-hPa) en i onmen al ela i e humidi y (“ hmd_ships”; %)
● Uppe - oposphe ic (500–300-hPa) en i onmen al ela i e humidi y (“ hhi_ships”; %)
● Dis ance o nea es majo landmass (“d l_ships”; km)
● Maximum po en ial in ensi y (“mpi_ships”; k )
● Reynolds sea su ace empe a u e (“ss _ships”; °C)
● Oceanic hea con en om he NCODA analysis (“ohc_ships”; J kg-1 °C-1)
● Mean o al p ecipi able wa e wi hin 200 km o he TC (“pw2m_ships”; mm)
● S anda d de ia ion o o al p ecipi able wa e wi hin 200 km o he TC (“pw2s_ships”; mm)
● Mean o al p ecipi able wa e wi hin 500 km o upshea quad an (“pw5u_ships”; mm)
● Mean o al p ecipi able wa e wi hin 500 km o he TC (“pw5m_ships”; mm)
● Nea es bes - ack ime (e.g., “yea _ships” o “hou _ships”)
All o he abo e pa ame e s a e a ailable in six-hou inc emen s, spanning 48 h p io o he ime o he
obse a ion, o 48 h ollowing he obse a ion (“ships_lag_ imes”), whene e analyses we e a ailable.
Please e e o SHIPS de elopmen al da a o mo e in o ma ion on a iable desc ip ions:
h p:// ammb.ci a.colos a e.edu/ esea ch/ opical_cyclones/ships/de elopmen al_da a.asp
5. Ve sion upda es:
3a: Implemen ed a be a e sion o he new cen e - inding me hod
3b: Fixed some bugs associa ed wi h cen e - inding echnique
3c: Adjus ed uning pa ame e s associa ed wi h cen e - inding algo i hm and imp o ed code e iciency
3d: Fixed a bug ha a ec ed he bes - ack en ies o s o ms wi h an analysis hou be ween 21–00
UTC. Replaced olde Ea l (2010) analyses wi h an upda ed QC e sion. Ve ical eloci y analyses o
s o ms p io o 2010 should s ill be used wi h ex eme cau ion (o a oided al oge he ).
3e: Added he capabili y o compu e o ex il om swa h da a. Imposed s ic e co e age c i e ia o
ecen e TC. I c i e ia a e no me , alues a e se o “NaN”.
3 : Added 2020 cases and implemen ed a e ised cen e - inding algo i hm, which is no longe
dependen on swa h RMWs. This change was de e mined o p o ide cen e es ima es ha be e ag eed
wi h subjec i e analyses.
3g: Added he beginning and end imes o swa h analyses. Also ixed a bug ela ed o he da e s amp
(da e ime objec s) wi h ligh s ha spanned bo h he las day o one mon h and he i s day o he nex
mon h (e.g., e ening/o e nigh ligh s on Augus 31). Al hough his bug only a ec ed a ew cases, he
co esponding me ged analyses we e inco ec ly unable o be pai ed wi h bes - ack and SHIPS
me ada a.
3h: Co ec ed a bug in he la i udes and longi udes o each TDR swa h and me ged analysis. The main
modi ica ion was o co ec ly accoun o he cu a u e o he Ea h. The e was also an adjus men o
accoun o he ac ha he o iginal TDR analysis g id o igin alls be ween g id poin s, a he han a he
middle g id poin . Due o hese changes, he loca ion o some TC cen e s/ il s we e changed in his
e sion.
3i (Used in TC-RADAR pape ): Co ec ed a bug in he “ge _bea ing” ou ine ha compu es he angle o
each g id poin . This co ec ion a ec ed he cen e - inding algo i hm. Cen e s seem la gely simila o he
p e ious e sion ( 3h), bu his me hod is mo e accu a e.
3j (5/25/2022): Added 2021 cases. Also implemen ed a e ised cen e - inding algo i hm designed o
mo e accu a ely es ima e TC cen e es ima es om swa h da a, especially in he mid-le els o weak TCs.
This included an addi ional co e age cons ain , which is mo e s ic han he p e ious e sion ( 3i).
Ve sion 3j also includes a bug ix o he calcula ion o di e gence.
3k (6/22/2023): Added 2022 cases. Sligh ly modi ied he cen e - inding me hod o use a la ge “co e
adius” pa ame e o imp o e he consis ency o he me hod on swa h analyses. As new analyses ha e
been added o he da abase, he size o he .nc iles ha e also g own. To mi iga e memo y issues, he
TC-RADAR iles ha e been sepa a ed in o wo swa h analysis iles and wo me ged analysis iles, g ouped
by consecu i e yea s (e.g., 1997–2019 and 2020–2022). Some coo dina e name lis pa ame e s we e
e ised o be mo e use iendly.
3l (10/10/2024): Added 2023 cases. Re ised names o a iables o be mo e consis en wi h ne CDF CF
con en ions. Also added “long names” o a iables. Las ly, ou p ecipi able wa e a iables we e added
om he SHIPS de elopmen al da a.
3m (8/26/2025): Added 2024 cases. Added nea es bes - ack ime in o ma ion in o he g oup o SHIPS
a iables o easie ime e e encing and o aid po en ial in e pola ion o en i onmen al da a o he ime
o TDR analysis. The cen e - inding me hod was op imized such ha a c ude ini ial sea ch was no longe
needed. This change should no a ec he la ge majo i y o cen e es ima es, bu may sligh ly enhance
he accu acy o swa h il es ima es o a hand ul o cases. Ou side o changes o he sea ch pa ame e s,
he cen e - inding weigh s and cos unc ion we e unchanged.
6. Da a access/Usage policy:
Each TC-RADAR ne CDF ile can be downloaded om he ollowing loca ion:
h ps://www.aoml.noaa.go / p/pub/h d/da a/ ada /le el3/. Use s a e ee o download TC-RADAR,
howe e , we eques use s no i y he HRD ada eam lead, D . Paul Reaso (email:
paul. e[email p o ec ed]), o he na u e o he esea ch p ojec o ensu e p ope usage. The na u e o he
quali y con ol (QC) p ocess is a delica e balancing ac be ween e aining as much me eo ological da a as
possible, while emo ing non-me eo ological a i ac s. Mo e speci ically, while he QC o aw ada da a
(Le el 1a) a emp s o iden i y and emo e non-me eo ological da a, some me eo ological da a is
in a iably emo ed in he p ocess. We ha e de e mined ha o many esea ch applica ions maximizing
co e age is essen ial. The e o e, he QC se ings employed he e do no aim o emo e all
non-me eo ological da a in he p oduc ion o he Le el 2 ada analyses con ained wi hin TC-RADAR.
De ails on he p oduc ion o speci ic Le el 2 analyses since 2020 a e p o ided in he ollowing ada da a
README: h ps://www.aoml.noaa.go /wp-con en /uploads/2023/04/README_ ada _da ase .pd . I is
incumben on he use o ca e ully examine he wind and e lec i i y analyses and employ an addi ional
le el o QC o he g idded ields consis en wi h hei objec i es. Use s also should be awa e o he
known issues documen ed in Sec ion 7. Any ollow-up ques ions can be di ec ed o he HRD ada eam.
7. Known issues:
The ollowing a e known issues wi h he da abase ha use s need be awa e o :
● TDR analyses o s o ms p io o he 2010 hu icane season we e c ea ed wi h an e o in he
syn hesis so wa e ha esul ed in e ical eloci ies ha we e oo deep and oo s ong in some
ins ances. As such, quan i a i e analyses o he TDR-de i ed e ical eloci y ield should no
be made o analyses p io o 2010. Analyses since 2010 (inclusi e) ha e add essed his e o
and a e conside ed mo e accu a e. This issue is documen ed in mo e de ail in Fische e al.
(2022).
● TDR-de i ed ada e lec i i y has no ye been p ope ly calib a ed. Re lec i i ies can a y
signi ican ly (some imes up o 7–10 dB) be ween he ada sys ems onboa d each NOAA P3 (e.g.,
N42 s. N43) in some seasons. The e a e also di e ences in e lec i i y ac oss seasons as he
ada sys ems onboa d each ai c a changed o e ime. Un il a p ope calib a ion is pe o med,
e lec i i y should no be used o quan i a i e analyses ou side o indi idual case s udies.
Because he TDR is an X-band ada , a enua ion also a ec s he analyzed e lec i i ies,
pa icula ly along he edges o swa hs in egions o hea y p ecipi a ion. The e ec s o
a enua ion ha e also no been accoun ed o in TC-RADAR.
8. How o ci e:
Any publica ions and/o p esen a ions using TC-RADAR should ci e he ollowing publica ion:
Fische , M. S., R. F. Roge s, P. D. Reaso , and J. F. Gamache, 2022: An analysis o opical cyclone
o ex and con ec i e cha ac e is ics in ela ion o s o m in ensi y using a no el ai bo ne
Dopple ada da abase. Mon. Wea. Re ., 150, 2255–2278,
h ps://doi.o g/10.1175/MWR-D-21-0223.1.
Any e e ence o he cen e - inding and/o o ex il should also e e ence:
Fische , M. S., P. D. Reaso , J. P. Dunion, and R. F. Roge s, 2024: An Obse a ional Analysis o he
Rela ionship be ween T opical Cyclone Vo ex Til , P ecipi a ion S uc u e, and In ensi y Change.
Mon. Wea. Re ., 152, 203–225, h ps://doi.o g/10.1175/MWR-D-23-0089.1.
9. Re e ences:
Gamache, J.F., F.D. Ma ks, and F. Roux, 1995: Compa ison o Th ee Ai bo ne Dopple Sampling
Techniques wi h Ai bo ne In Si u Wind Obse a ions in Hu icane Gus a (1990). J. A mos.
Oceanic
Technol., 12, 171–181, h ps://doi.o g/10.1175/1520-0426(1995)012<0171:COTADS>2.0.CO;2
Lo solo, S., J.A. Zhang, F. Ma ks, and J. Gamache, 2010: Es ima ion and Mapping o Hu icane Tu bulen
Ene gy Using Ai bo ne Dopple Measu emen s. Mon. Wea.
Re ., 138, 3656–3670, h ps://doi.o g/10.1175/2010MWR3183.1
Reaso , P.D., M.D. Eas in, and J.F. Gamache, 2009: Rapidly In ensi ying Hu icane Guille mo (1997). Pa I:
Low-Wa enumbe S uc u e and E olu ion. Mon. Wea.
Re ., 137, 603–631, h ps://doi.o g/10.1175/2008MWR2487.1
Reaso , P.D. and M.D. Eas in, 2012: Rapidly In ensi ying Hu icane Guille mo (1997). Pa II: Resilience in
Shea . Mon. Wea. Re ., 140, 425–444, h ps://doi.o g/10.1175/MWR-D-11-00080.1
Roge s, R., S. Lo solo, P. Reaso , J. Gamache, and F. Ma ks, 2012: Mul iscale Analysis o T opical Cyclone
Kinema ic S uc u e om Ai bo ne Dopple Rada Composi es. Mon. Wea.
Re ., 140, 77–99, h ps://doi.o g/10.1175/MWR-D-10-05075.1
