D ependent V erification in Italy Adriano Raspanti Maria Stefania Tesini summary Subjective classification at IMS COSMOMED verification against synop stations 2m Temperature 10 m Wind speed ID: 493107
Download Presentation The PPT/PDF document "W eather type" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Weather type Dependent Verificationin ItalyAdriano RaspantiMaria Stefania TesiniSlide2
summarySubjective classification at IMSCOSMO-MED verification against synop stations:2m Temperature10 m Wind speedCloud coverTotal Precipitation (6h)COSMO-MED,COSMO-I7,COSMO-I2,ECMWF against high resolution raingauges:Total Precipitation (24h)Slide3
summarySubjective classification at IMSCOSMO-MED verification against synop stations:2m Temperature10 m Wind speedCloud coverTotal Precipitation (6h)
COSMO-MED,COSMO-I7,COSMO-I2,ECMWF against high resolution raingauges
:Total Precipitation (24h)Slide4
Subjective Classification at IMS COSMO GM 2011 - ROMA
1
2
3
4
5
6
7
8
9
10
11Slide5
Subjective Classification at IMS
COSMO GM 2011 - ROMASlide6
summarySubjective classification at IMSCOSMO-MED verification against synop stations:2m Temperature
10 m Wind speed
Cloud coverTotal Precipitation (6h)
COSMO-MED,COSMO-I7,COSMO-I2,ECMWF against high resolution
raingauges
:
Total Precipitation (24h)Slide7
NCMCCML
CMT
Temperature
Negative bias for all cases. Improvement for CMT and CML in RMSE
All casesSlide8
MCCMLCMT
NC
WIND SPEED
No clear differences
All cases
COSMO GM 2011 - ROMASlide9
Total Cloud CoverCMLCMT
All cases
Less overestimation for CMT. Almost no difference in RMSE or MAESlide10
Total Cloud CoverNCMC
All cases
Improvement in bias for NC and worsening for MC. RMSE the same
COSMO GM 2011 - ROMASlide11
TP 06H FBINCMC
CML
CMT
All cases
Less overestimation for CML case and bias around 1 for many thresholds. Higher overestimation for NC
COSMO GM 2011 - ROMA
FBI = 1
FBI = 1
FBI = 1
FBI = 1
FBI = 1Slide12
TP 06H ETSAll casesNC
MC
CMT
CML
Worse or similar ETS for NC, MC and CMT. Slightly better for CML
COSMO GM 2011 - ROMASlide13
TP 06H FBINCMC
CML
CMT
All cases
Different behaviour for lower and higher thresholds
FBI = 1
FBI = 1
FBI = 1
FBI = 1
FBI = 1Slide14
TP 06H ETSAll casesNC
MC
CMT
CML
Worse or similar ETS for NC, MC and CMT. Slightly better for CML
COSMO GM 2011 - ROMASlide15
summarySubjective classification at IMSCOSMO-MED verification against synop stations:2m Temperature10 m Wind speedCloud coverTotal Precipitation (6h)
COSMO-MED,COSMO-I7,COSMO-I2,ECMWF against high resolution
raingauges
:
Total Precipitation (24h)Slide16
COSMO GM 2011 - ROMADay 1Day 2…Day nArea 1
Area 2
…
Area 96
S
Mean
/
Median
value
of
precipitation
For
each
weather
type
category
:
Daily
scores
Scores
for
the
selected
categorySlide17
2-Zonal Westerly
cyclonicSlide18
4-Meridional
cyclonicSlide19
6-Northerly
cyclonicSlide20
7-Northerly
anticyclonicSlide21
8-Central Mediterranean
HighSlide22
9-Central Mediterranean
RidgeSlide23
10-Central Mediterranean LowSlide24
11-Central Mediterranean
TroughSlide25
2ZWC4MC6NC7NA8
CMH9CMR
10CML11
CMT
ALL
WEST
ALPS
i
n
n
n
h
h
h
h
h
h
nh
hhni
hnnh
hhhhh
hhnh
hinh
hnREAST ALPS
nnh
nn
hh
nn
hnh
n
hin
in
hh
nh
h
h
n
n
n
n
h
n
i
n
n
R
R
R
NORTH
WEST
i
i
n
i
h
h
n
n
h
n
i
h
h
h
n
n
n
n
n
n
n
h
n
n
h
h
h
n
n
n
i
i
R
R
R
R
PO VALLEY
i
i
n
i
h
h
n
i
n
i
i
i
h
h
n
h
n
h
h
h
n
h
h
n
n
n
n
i
h
n
n
n
R
R
n
R
NORTH
APPEN
i
n
n
n
n
h
n
n
n
n
n
n
h
n
h
n
n
h
h
h
n
h
h
h
n
h
h
h
n
n
i
i
R
R
R
R
SOUTH
APPEN
i
i
h
i
i
i
i
i
i
n
i
n
n
i
n
i
i
i
i
i
i
n
i
n
i
n
h
h
n
n
n
n
i
R
R
R
TIRRENIAN COST
i
i
h
i
n
i
n
i
n
i
i
i
n
n
i
n
n
n
n
n
h
n
n
h
h
n
h
i
n
i
i
i
R
R
R
R
ADRIATIC COAST
i
n
n
i
n
n
n
n
n
h
n
n
h
n
n
n
n
n
n
n
n
n
n
n
n
i
nnniiiRnnnSOUTHnnnnniiihiinhiinnnnnnhnninhhniihnRRRGloballyiRnnRRRRRRRRhRRnnnnRRhRRRRhRnRiR
ECMWF COSMO-I7 COSMO-MED COSMO-I2
h
overestimation
i
understimation
n
almost
correct
R
variable
behaviourSlide26
Some considerations on the rough estimate of the amount of rainThe dataset does not cover equally all the territory so the results are just an indication It is very difficult to asses the behavior of models in a particular synoptic situation over all the italian region due to complex orography In each area models behave in a different way according to the synoptic situationSlide27
Small dots =
daily scoresBig dots
= scores over the days in each category
All
casesSlide28
Small dots =
daily scoresBig dots
= scores over the days in each category
All
casesSlide29
Small dots
= daily scoresBig
dots = scores over the days in each category
All
casesSlide30
Some considerations on models performances At low threshold (e.g. 1 mm/24h) Cosmo Models perform well in cyclonic situations (CLM,CMT,MC) – high TS and BIAS ≈1but some difficulties in NC ECMWF is strongly biasedIn anticyclonic situation COSMO-MED and ECMWF are better in terms of POD but they tend to overestimate the number of eventsAt higher thresholds (e.g. 5 m/24h and 10 mm/24h)COSMO-I7 and I2 miss the anticyclonic situation (except MA ) still good performance for all models for the cyclonic situation (except for NC)ECMWF reduces the BIAS SCORE Note the different scores for each day of a selected category!!Slide31
ConclusionSynoptic verification of COSMO-MED did not point out significant differences between the selected categoryHigh resolution verification showed differences in the behavior of models over italian regions, according to weather type category, but the results are difficult to interpret The good news is that models are able to reproduce more or less all the type of weather!! Slide32
Thanks for your attention