Content uploaded by Agostino Niyonkuru Meroni
Author content
All content in this area was uploaded by Agostino Niyonkuru Meroni on Jun 19, 2019
Content may be subject to copyright.
ASSIMILATION OF SENTINEL-DERIVED AND GNSS-DERIVED
PRODUCTS IN HIGH IMPACT WEATHER EVENTS NUMERICAL
SIMULATIONS: THE ESA-STEAM PROJECT RESULTS
M . La gas i o 1 (ma r ti na .la g as io @ci m af ou nda t io n. org ) , A. Pa r od i 1, L . Pu lvi r en ti 1, G. B oni 2, N . P i er di cca 3,
G . Ve nut i 4 , A. N . Me ron i 1 ,4 , E . R ea lin i 5 , A. G at ti 5, S. B ari n de ll i 4 , B . Rom m en 6
MOT IVA T ION
EAR TH O BSE RVA T ION PRO DUC T S A ND M ODE L S E TUP
The re is a stron g int erest in
for ecast ing heav y rai nfall events
bec ause of their high impa ct on th e
soc iet y
Num erica l weathe r pre dicti on
mod els s uffer fr om co arse resoluti on
ini tial and boun dary condi tions
RES ULT S
< W RF do mains V3.8. 1 at 13.5, 4.5 and 1 .5 km .
Ini tial and bo undar y con ditio ns co me fr om bo th
NCE P-GFS and E CMWF- IFS.
Pro ducts to be a ssimi lated :
Soi l Moi sture, S M (nu dging -like)
Sur face wind, WI ND (3 DVAR)
Sea Surf ace Temp eratu re, S ST (dire ct in sertion)
Lan d Sur face Tem perat ure, LST (dir ect
ins ert ion)
Zen ith T otal Del ay, Z TD, f rom GNSS (3DV AR)
Zen ith T otal Del ay fr om In SAR (3DV AR)
> ( a) So il moi sture
[m^ 3/m^3 ] 18 U TC
08/ 10/20 17 (b)
Sur face wind f ield
[m/ s] 18 UTC
08/ 10/20 17 (c) Sea
sur face temper ature
[K] 21 U TC
09/ 10/20 17 (d) Land
sur face temper ature
[K] 10 U TC
09/ 10/2 017.
Val idati on is pe rform ed wi th the M ODE t echnique [Dav is
et al, 2 006a,b] that calcu lates so me sp atial in dices to
com pare the fore caste d fie ld and t he ob served o ne.
The timi ng and s patia l cov erage of the satellit es ar e
con strai ned by t he sa telli te swat and, thus, ar e not
alw ays t he most suita ble: more exp erime nts on t hat a re
pla nne d.
1 C IM A R e se ar ch F ou nd ati o n
2 U ni ver s it y of G en oa
3 S ap ien z a Un ive r si ty of Ro me
4 P ol ite c ni co di Mi la no
5 G eo mat i cs R ese a rc h and
De vel o pm en t
6 E SA -ES T EC
RESEARCH QUESTION
Does the assimilation of high
resolution Earth Observation
(EO) variables improve the
forecast of heavy rain events?
Var ious data ass imila tion experime nts a re
per forme d for tw o cas e stu dies:
the Livo rno 9-10 Sept ember 2017 ca se (b rief and
loc ali zed)
the Silv i Marina 14-1 5 Nov ember 20 17 ca se
(ex tende d and lo ng la sting )
[Mo lini et al, 2 011]
^ Z TD [c m] fro m InS AR at 05 U TC 14 /11/2 017,
ove rlaid with the G NSS-d erive d ZTD .
TAKE HOME MESSAGE
Assimilating satellite-derived
wind and/or water vapor
content appears to have the
largest positive impact on the
forecast of heavy rain events.
< S ILVI MARINA
a=S ST, b =SM-Se ntine l 1,
c=S M-SMA P/Sent inel 1,
d=S M-SMA P, e=W IND,
f=Z TD-I NSAR ,
g=Z TD3h _1is t,
h=W IND+ SM+I NSAR+ GN
SS_ only 5UTC ,
i=O bserv ations , j=O pen
Loo p
> L IVORN O
a=L ST, b =SST, c=SM,
d=W IND, e=ZTD3 h,
f=Z TD3h _1is t,
g=W IND+ SM+Z TD,
h=W IND+ SM+Z TD_on ly1
8UT C, i= Observ ation s,
j=O pen L oop
^ M ap of the G NSS r eceiv ers
ass imila ted in the exper iment s.
The sate llite pr oduct s are assimil ated at the t ime
of the p assage. The Z TD fr om GNSS is as similate d
eve ry th ree hour s (ex cept when exp licit ly
wri tte n).
3DV AR is perform ed wi th WR FDA [Bar ker, et al,
201 2] V3 .9.1. Th e bac kgrou nd error cova riance
mat rix i s estima ted w ith t he Natio nal
Met eorol ogical C enter meth od.
REF EREN CES
Bar ker e t al (2012 ), Bu ll. Amer. Mete or. S oc., 93, 831–8 43.
Mol ini e t al (2011 ), Q. J. Royal Mete orol. Soc ., 13 7(654 ), 14 8-15 4.
Dav is et al ( 2006a ), Mo n. W eathe r Rev ., 13 4, 1 772–1 784.
Dav is et al ( 2006b ), Mo n. W eathe r Rev ., 13 4, 1 785–1 795.