COSMO General Meeting Five Applications · 2016. 12. 20. · Jacques Ambühl, Meteo Swiss; COSMO GM...
Transcript of COSMO General Meeting Five Applications · 2016. 12. 20. · Jacques Ambühl, Meteo Swiss; COSMO GM...
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
COSMO General Meeting
Five Applications
Sibiu, 2-6 Sept 2013Jacques Ambühl
Samstag, 31. August 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Plan
Use of COSMO-7 data to generate power production forecasts for photovoltaic plantsTowards operational Clear Air Turbulence diagnostics with COSMO-7Dispersion forecast for nuclear pollutantsGale warnings with genetic algorithms using COSMO-2 dataOptimization of sequential sailing decisions based on COSMO-2
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Use of COSMO data to generate power production forecasts for photovoltaic plants
Dominik Büeler ETHZ IAC
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Generate day-ahead power production forecasts for a Swiss wide system of photovoltaic (PV) plants in 15-min resolution based on COSMOBenefit of COSMO (mainly irradiance data) in meeting the needs of the energy sector for highest-resolution weather prediction data both temporally and spatiallyQuantification and discussion of meteorological and technical errors / uncertainties associated with a PV power production forecast based on COSMO
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Data Source Temporal resolution Spatialresolution
COSMO-2 MeteoSwiss 1 h 2.2 x 2.2 km
Satellite MeteoSwiss 15 min / 24 h 1.1 x 1.7 km
Surface observations MeteoSwiss 10 min / 24 h SwissMetNet Stations
PV metadata ewz - PV plants
PV production data ewz 15 min PV plants
Balancing energy prices Swissgrid 15 min Switzerland
COSMO-2: most representative grid point for each PV plant location as data basis Satellite: closest grid point for each PV plant location as data basis
All data are available for the years 2010 – 2012
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Method «Tilted»: direct and diffuse irradiance on a horizontal plane (from COSMO-2) are transformed on a tilted and oriented PV surface and summed up to the global irradiance
S
W
21°Idiffuse-tilted
Idirect-tilted
Iglobal-tilted+
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
PV forecast verification: «Tilted» | time series example | March 3 – 6, 2012
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
PV forecast verification: «Tilted» | RMSE | 2012 | corrected (only for prod. > 0.5 kWh)
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Towards operational Clear Air Turbulence diagnostics with COSMO-7
Fabian Binder ETHZ IAC
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
CAT Indexes:- Gradient Richardson Number RI- Ellrod & Knapp Indexes TI1 & TI2- Trend Term DTV- Divergence Modified Turbulence Index DTI ∋ {TI1, DTV- Turbulent Kinetic Energy (TKE)
Opt. Thresholds- RI: 0.7; TI2: 4.0E-6 [s-2]; TKE: 0.6 [J]; DTI: 7.0E-6 [s-2]
Trade off between thresholds and Volume event Tests in 2014
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
14
a)
b) c)
Figure 7: The encounter location is shown as cross (+) in a) and c) respectively as intersecion of alongwind (red line) and crosswind (green line) cross section in b). a) shows the vertical cross section in alongwind direction of TKE, TH (contours) and 2pvu (thick line). a) and b) are horizontal cross sections at flight altitude whereas b) is a zoom into the closer region.
met
er
TKE − alongwind cross section 21.01.2011 10 UTC
−6 −5.8 −5.6 −5.4 −5.2 −5 −4.8 −4.6 −4.4 −4.20
2000
4000
6000
8000
10000
12000
00.20.40.60.811.21.41.61.822.22.42.62.833.23.43.63.844.24.44.64.855.25.45.65.8
TKE @ 10910 m flight altitude 21.01.2011 10 UTC
−15 −10 −5 0 5
−8
−6
−4
−2
0
2
4
6
8
10
00.20.40.60.811.21.41.61.822.22.42.62.833.23.43.63.844.24.44.64.855.25.45.65.8
TKE @ 10910 m flight altitude (zoom) 21.01.2011 10 UTC
−6 −5.8 −5.6 −5.4 −5.2 −5 −4.8 −4.6 −4.4 −4.2
−3.2
−3
−2.8
−2.6
−2.4
−2.2
−2
−1.8
−1.6
−1.4
00.20.40.60.811.21.41.61.822.22.42.62.833.23.43.63.844.24.44.64.855.25.45.65.8
Example CAT Event 21.01.2011: Massif Central (F):Mountain Gravity Wave
Source: Swiss Airlines, Flight Data Management
Outcome: objective warning tool for aviation forecasters
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Dispersion forecast for nuclear pollutants
Pirmin KaufmannMeteoSwiss, Federal Administration
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
COSMO-7Dispersion path of an air parcel starting from ZürichFast estimation
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
COSMO-2Dispersion & concentration of radioactive pollutants starting from Gösgen.Comprehensive simulation
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Gale warnings with genetic algorithms using COSMO-2 data
Roman Voisard, Manuela Züger Uni ZH AI Lab Lysiane Mayoraz ETHZ IAC
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz!"! #$%$&'()$*+,'-,.,/+'0),1'0$2.3+,14*2+5'*,65+7,89!
,
,
15:40$,;<,=4+').+52,$.0+7,3+.+5'*3,'-,/6533>$+?$+,*$.0,80$*27$*,@>$+$'/6533A,BC!CD,
10'), $.27, '-, +7$3$, 3+.+5'*3, ���� ������ �� ����������� ���� ������ �� ���� ���� ������ ���5)(&$)$*+.+5'*E,F7$3$,(.0.)$+$03,.0$,)$.340$G,$%$0H,+$*,)5*4+$3E,
I.05.J&$, #$3205(+5'*, K*5+,GL&C!CMC,,, N5*G,G50$2+5'*O,+$*,)5*4+$3,)$.*,, #$:0$$,(0$3+.3C,,, =50, (0$3340$, .+, 3+.+5'*, .&+5+4G$, @P1QDO, 5*3+.*+.*$'43,
%.&4$,R$2+',9.32.&,
-4BC!CM!,, 9$.L,'-,:43+,@3$2'*GDO,).S5)4), T*'+,@*.4+E)5&$,(E7'40D,-4BC!CMC,,, N5*G,3($$GA,+$*,)5*4+$3,)$.*, T*'+,@*.4+E)5&$,(E7'40D,+0$BCC3C,,, =50,+$)($0.+40$,B),'%$0,+$00.5*O,5*3+.*+.*$'43,%.&4$, #$:0$$,U$&3543,+G$BCC3C,,, #$6,('5*+,B),'%$0,+$00.5*A,5*3+.*+.*$'43,%.&4$, #$:0$$,U$&3543,
F.J&$,!<,9.0.)$+$03,(0'%5G$G,JH,/6533>$+?$+,3+.+5'*3,
V*&H, G.+., 0$2'0G3, *'+, 2'*+.5*5*:,?=?W%.&4$3, @*', *4)J$0, :5%$*, -'0, ., 3($25-52, G.+$, .*G, +5)$D, .0$,2'*35G$0$GE,
"#$%&$#'()!*#+,#-$(.,
F', 37.0($*, +7$, G.+.A, G5--$0$*2$3, '-, (.0.)$+$03, .0$, +.L$*, 5*3+$.G, '-, +7$, 0.6, %.&4$3E, F753,).L$3,3$*3$,-'0,+7$,(.0.)$+$03,(0$3+.3C,@.50,(0$3340$D,.*G,+0$BCC3C,@.50,+$)($0.+40$DE,U'40%'535$0,-'4*G,$%5G$*2$,-'0,.,2'00$&.+5'*,J$+6$$*,+$)($0.+40$,:0.G5$*+3,.*G,65*G,3+0$*:+7,.*G,J$+6$$*,(0$3340$,:0.G5$*+3, .*G, 65*G, 3+0$*:+7, 0$3($2+5%$&H, @U'40%'535$0A, !;"XDE, #5--$0$*2$3, .0$, +.L$*, J$+6$$*,3+.+5'*3, 67527, &5$, 5*, '(('35+$, G50$2+5'*3, .*G, 7.%$, 80$*27$*,)'0$, '0, &$33, .3, +7$50, )5GG&$, ('5*+E,1'&&'65*:,.GG5+5'*.&,%.05.J&$3,.0$,2.&24&.+$G<,
!"#$%#&'()*+$)",)-$%$#&.)!/"0/'11&%0)&%)2#"/1)3'/%&%0)24+#$1+) !")#
#
$%&'()#*+#,-./01)#23#4'5#6())#7(24428)(#9:21%;#<.=>2=;#?#@7:A));#BCCDE#
!"#$#%&'#%4#6A)#7().6%2=#23#.#=)F#7A%1>#0(2&(./#5G#(.=>2/1G#.16)(%=&#.#(.=>2/1G#7A24)=#0.(6#23#.#
4)1)76)>#0.()=6#0(2&(./# 9:21%;# <.=>2=;#?#@7:A));#BCCDEH# I6# 72=4%464# %=# 7A.=&%=&# 6A)#72=6)=6#23#.#
(.=>2/1G# 4)1)76)># =2>)# F%6A# .=26A)(# )1)/)=6# 3(2/# 6A)# 4./)# 0(%/%6%8)# 4)6H# $%&'()# D# 4A2F4# .=#
)-./01)#23#4'5#6())#/'6.6%2=H#
()*+&,"-#%&'#%4#6A)#7().6%2=#23#.#7A%1>#0(2&(./#FA%7A#%4#.=#)-.76#720G#23#.#4)1)76)>#%=>%8%>'.1#23#6A)#
0()8%2'4# &)=)(.6%2=# 9:21%;# <.=>2=;# ?# @7:A));# BCCDEH# I=# JKL:;# 6A)# 0(25.5%1%6G# 23# 7(24428)(# .=>#
()0(2>'76%2=#/'46#.>>#'0#62#!H#
#
$%&'()#D+#,-./01)#23#4'5#6())#/'6.6%2=#9:21%;#<.=>2=;#?#@7:A));#BCCDE#
!"#"$ %&'()�)'*)+,-*)./012-)&-*-342)'56/143789):1;8-</1=)
300 - 1000 generations1000 individuals per generationensembles with 20 individuals
Herds dynamics: nowcasting of gales with genetic programming
System under further development
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
divide.evaluate(pow.evaluate(fmo, 7.598043337179606), dmo)
Depth = 3Fitness = 5487.260577858043Hit Rate = 98.25662482566248FalseAlarmratio = 44.15378517637733TestFitness = 5450.160304096759Test Hit Rate = 97.48603351955308TestFalseAlarmratio = 44.092911493792556
plus.evaluate(5.295243401974644, exp.evaluate(minus.evaluate(pow.evaluate(8.235782453080361, min.evaluate(sine.evaluate(tt40), 7.224975929882489)), exp.evaluate(sine.evaluate(multiply.evaluate(log.evaluate(pow.evaluate(log.evaluate(5.426030439601154), exp.evaluate(2.016302169365234))), minus.evaluate(1.5110163068855542, tt60)))))))
Depth = 10Fitness = 803.8617518559158Hit Rate = 13.94700139470014FalseAlarmratio = 42.363112391930834TestFitness = 431.4222138055109Test Hit Rate = 9.427374301675979TestFalseAlarmratio = 54.23728813559322
Herds dynamics: nowcasting of gales with genetic programming
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
divide.evaluate(pow.evaluate(plus.evaluate(divide.evaluate(dmo, tt20), f00), max.evaluate(pow.evaluate(2.828094360716278, plus.evaluate(tt20, divide.evaluate(tp40, d20))), d60)), plus.evaluate(tp60, ifElseThen.evaluate(divide.evaluate(1.5081885565933717, d40), max.evaluate(minus.evaluate(minus.evaluate(ifElseThen.evaluate(f20, d00, multiply.evaluate(plus.evaluate(sine.evaluate(divide.evaluate(log.evaluate(ifElseThen.evaluate(tp40, minus.evaluate(f00, d60), 5.7217590097084, 4.168310361526711)), pow.evaluate(max.evaluate(plus.evaluate(multiply.evaluate(multiply.evaluate(pow.evaluate(minus.evaluate(f20, plus.evaluate(qf40, 7.074758580356248)), 3.9596438991780456), max.evaluate(exp.evaluate(tt40), mmo)), multiply.evaluate(tp40, d20)), ifElseThen.evaluate(3.5117337093986056, pow.evaluate(max.evaluate(min.evaluate(f00, plus.evaluate(7.0407142440130235, tp40)), log.evaluate(minus.evaluate(8.215013810404919, 5.686086056525813))), ifElseThen.evaluate(minus.evaluate(tp60, exp.evaluate(f00)), max.evaluate(fmo, 2.077248774619843), ifElseThen.evaluate(qf40, ifElseThen.evaluate(d60, f40, tt40, tt60), 9.819823338454908, ifElseThen.evaluate(0.1293375089801585, 6.171938941943528, tp60, tpt)), ttt)), min.evaluate(sine.evaluate(sine.evaluate(min.evaluate(4.356468293326104, 3.7643533991009437))), log.evaluate(8.924252353471719)), tpt)), qf60), minus.evaluate(divide.evaluate(pow.evaluate(sine.evaluate(max.evaluate(sine.evaluate(plus.evaluate(f60, 4.901973747799307)), multiply.evaluate(d60, 4.444514844502521))), 2.8856566954867113), ifElseThen.evaluate(3.5282325518722955, divide.evaluate(min.evaluate(d00, qf60), qf40), 2.782942429981783, ttt)), minus.evaluate(1.5471842086219156, 1.4214887917363217))))), minus.evaluate(exp.evaluate(f00), min.evaluate(d60, max.evaluate(max.evaluate(plus.evaluate(multiply.evaluate(log.evaluate(sine.evaluate(plus.evaluate(qf40, tt20))), multiply.evaluate(tp60, qf40)), minus.evaluate(tt60, 7.393229781243173)), max.evaluate(multiply.evaluate(sine.evaluate(log.evaluate(mmo)), log.evaluate(log.evaluate(min.evaluate(divide.evaluate(f20, 2.5355487159252474), exp.evaluate(tp60))))), d20)), mmo)))), max.evaluate(dmo, d20)), log.evaluate(min.evaluate(plus.evaluate(ifElseThen.evaluate(qf20, tp40, minus.evaluate(fmo, divide.evaluate(mmo, fmo)), max.evaluate(dmo, mmo)), 4.703888157585869), max.evaluate(6.941210952688398, ifElseThen.evaluate(multiply.evaluate(6.1221500448510655, min.evaluate(plus.evaluate(1.9720986506835103, 2.4887295465961987), ifElseThen.evaluate(exp.evaluate(multiply.evaluate(0.7940071865385323, qfe)), plus.evaluate(d40, qf20), sine.evaluate(7.759870749566222), multiply.evaluate(plus.evaluate(1.6441563478507293, f60), tt40)))), f60, multiply.evaluate(tt40, plus.evaluate(multiply.evaluate(plus.evaluate(exp.evaluate(ifElseThen.evaluate(max.evaluate(qfe, log.evaluate(2.795088277634706)), qf20, ifElseThen.evaluate(qf40, sine.evaluate(0.5427474597208304), divide.evaluate(4.891146685200969, tp40), divide.evaluate(9.211053729636866, d20)), 3.3613880147278485)), max.evaluate(minus.evaluate(2.1476839510359813, 6.224490292998265), min.evaluate(ifElseThen.evaluate(tp20, pow.evaluate(6.055036006277885, 5.624835015943659), min.evaluate(1.2461272501542653, 9.388307443376117), f20), pow.evaluate(pow.evaluate(tp20, d60), sine.evaluate(5.278444724215593))))), f00), sine.evaluate(ttt))), qfe))))), min.evaluate(qfe, 4.608298012209561)), min.evaluate(min.evaluate(fmo, divide.evaluate(log.evaluate(max.evaluate(tt60, tp60)), 1.9858918908009104)), log.evaluate(pow.evaluate(log.evaluate(max.evaluate(8.79562243801495, dmo)), ifElseThen.evaluate(d60, tp20, qf20, 5.589801634828774))))), qfe), 9.036110890239984, pow.evaluate(6.12076610093759, 4.5673174527558995))))Depth = 20Fitness = 5045.881402259277Hit Rate = 99.86052998605301FalseAlarmratio = 49.47071277346507TestFitness = 5042.2535211267605Test Hit Rate = 100.0TestFalseAlarmratio = 49.57746478873239
Herds dynamics: nowcasting of gales with genetic programming
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Herds dynamics: nowcasting of gales with genetic programming
Left: Kloten, without COSMO (only Observations)Right: Grenchen, with COSMO (FieldExtra) and Obs
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
0.2 0.4 0.6 0.8 1
0.2
0.4
0.6
0.8
112 Ø Met. Threshold
5.10.15.20.25.30.
35.
40.
45.
50.
55.
60.
65.
70.
75.
80.
85.
90.
95.
B_Simul_Ens_Exhib_29.05.09.nb 1
Efficiency of thewarning system
Customerexposure
OptimalWarning
ProbabilityCustomer profile
Warning profile (ROC)
Hit Rate
False Alarm
Motivation: aus der Finanzwelt !
Return(%)
Risk (volatility)
Low riskLow return
Medium riskMedium return
High riskHigh return
Inefficient portfolios& securities
Capital asset pricing
ROC vs Efficient Frontier
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Optimization of sequential sailing decisions based on COSMO-2 forecasts
Jacques Ambühl
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Optimization of weather dependent sequential processes
Seeking forward for options: building a tree of potential decisions
Following the best sequence backward from the objective, as provided by the tree
Sequential Decision: Dynamical Programming
StartObjective
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Out[322]=
15'
18'
21'
24'
27'
30'
33'
12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57'Initialization ê Forecast HUTCL: 0 ê 14COSMO-2 W_10M ê 2013-06-15
15'
18'
21'
24'
27'
30'
33'
12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57'Initialization ê Forecast HUTCL: 0 ê 14COSMO-2 W_10M ê 2013-06-15
15'
18'
21'
24'
27'
30'
33'
12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57'Initialization ê Forecast HUTCL: 0 ê 23COSMO-2 W_10M ê 2013-06-1512'
15'
18'
21'
24'
27'
30'
33'
9' 12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57' 0'
Réseau de Décision
Right:COSMO-2 10m Windfield (Extraction: FieldExtra)
Below:Decision network(with departure and arrival nodes)
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Out[307]=
15'
18'
21'
24'
27'
30'
33'
12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57'
15'
18'
21'
24'
27'
30'
33'
12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57'
©Routage 2011: J. Ambühl
Date: 2013-06-15
Forecasting Model: COSMO-2
Forecasting Range: 0 - 33 UTC
Departure @UTCD: @ 8h 0mm D
9
10 11
12 1314
1516 17
15'
18'
21'
24'
27'
30'
33'
12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57'
©Routage 2011: J. Ambühl
Date: 2013-06-15
Forecasting Model: COSMO-2
Forecasting Range: 0 - 33 UTC
Departure @UTCD: @ 8h 0mm D
9
10 11
12 1314
1516 17
Out[18]=12345678910111213141516
12345678910111213141516
30
60
90
120
150
Close Hauled
Beam Reaching
Beating
On Broad Reach
†W § =10 Kt Incidence 70°
Boat speed: 7.3 Kt
12345678910111213141516
12345678910111213141516
30
60
90
120
150
Close Hauled
Beam Reaching
Beating
On Broad Reach
†W § =20 Kt
Incidence 140°
Boat speed: 13. Kt
Right: Decision tree and optimal routeBelow: Polar diagram of the sailing boat(customer profile)
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Out[101]=
15'
18'
21'
24'
27'
30'
33'
12' 15' 18' 21' 24' 27' 30' 33' 36' 39' 42' 45' 48' 51' 54' 57'
171819
2021
2223
24
25
26
27
28©Routage 2011: J. Ambühl
Date: 2013-06-15
Yacht: S850_Proto
Forecasting Model: COSMO-2
Forecasting Range: 0 - 33 UTC
Departure @UTCD: @ 15h 49mm D
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; ECAM2011
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Classical technique in financial & energy tradingAlinghi, America Cup (EPFL, Prof. Dalang)
StartObjective
Seeking forward for options: building the tree
Following the best sequence backward from the objective
Sequential Decision: Dynamical Programming
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
Conclusions
The Customer profile is the yardstick in any optimization processin terms of her/his/its outcomeand exposure (risk profile)
Three fields of competence are merged Probabilistic forecastBio-inspired artificial intelligenceFinancial inspiration
“Computer to computer” serviceMeteoSwiss, COSMO applications: 110 Gbytes data / 24 hours
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
15
4.2 Jet Stream
Event 33 26.Aug 16:58 UTC Flight altitude 8147.9 m Location Southern UK
Figure 8: Horizontal cross section of horizontal windspeed and wind vectors at flight altitude. The location of the encounter is shown as black cross over southern UK. The red line locates the vertical cross section shown in figure 8.a), which is oriented normal to the wind direction.
Figure 8 shows the location of the encounter at the edge of a strong jet stream over
southern UK. Figure 9.b) shows that the encounter happened near an extratropical cyclone, and behind an occluded front. This event is located near several weather features such as rain clouds, front and jet stream. There is rainfall but no clouds reach the height of the location of the encounter (figs 9.c) and d)). This event does not show signs of disturbance of the horizontal flow field caused by convection, as seen in in the next case study 4.3 Deep Convection (fig 12.c)).
The incident happened at the exit of a strongly curved jet streak (fig 8) at the height of the tropopause (fig 9.a)). The encounter is located at the edge of the jet, slightly below the height of maximum windspeed, in a region with horizontal and vertical windshear.
This event is detected by all CAT indices. Figures 10.a) and c) show that the encounter is precisely located by TI2 in the horizontal as well as in the vertical.
Wind speed [m/s], wind vectors, 2pvu @ 8147.9m (flight altitude) 26.08.2011 17 UTC
−15 −10 −5 0 5
−8
−6
−4
−2
0
2
4
6
8
10
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
16
a) b)
c) d)
Figure 9: The location of the encounter is shown as cross (+). a) shows the horizontal wind speed, TH (contours) and 2pvu (thick line), b) shows the synoptic ground chart (DWD, n.d.), c) shows the accumulated total rainfall and d) the vertical crossection in alongwind direction of TH, T (contours) and 2pvu (thick line).
met
er
Vertical Cross Section of Windspeed [m/s],Theta[K] and 2pvu 26.08.2011 17 UTC
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
222
2
22
2
22
2
290290
290
290
295295
295295 295
300 300
300300
300
305
305305
305
305
310
310310
310
310
315315
315
315
315
320 320
320
320
320
325 325325
325
325
330 330 330330
330
335 335335
335
335
340340
340
340
340
345345
345
345
345
350350
350
350
350
355355
355
355
355
360360
360
360
360365
365365365
370
370370370
375375375380
−10 −9 −8 −7 −6 −5 −40
2000
4000
6000
8000
10000
12000
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
Accumulated Total Rainfall [kg/m2] 26.08.2011 17 UTC
−15 −10 −5 0 5
−8
−6
−4
−2
0
2
4
6
8
10
024681012141618202224262830323436384042
met
er
Vertical Cross Section of RH and T 26.08.2011 17 UTC
−45−45 −45
−45
−45
−45
−45
−40
−40
−40−40
−40 −40−40
−35 −35−35
−35−35
−30 −30 −30−30 −30
−25 −25 −25
−25 −25
−20 −20 −20−20 −20
−15 −15 −15−15 −15
−10 −10 −10 −10−10
−5 −5 −5 −5 −5
0 0 0 0 0
5 5 5 5 510
10 10 10 1015
2
2 2 2
2
2
2
2
2
2
22
2
2
2
2
22
2
−7.2 −7.1 −7 −6.9 −6.8 −6.7 −6.6 −6.5 −6.4 −6.3 −6.20
2000
4000
6000
8000
10000
12000
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
Samstag, 31. August 2013
Jacques Ambühl, Meteo Swiss; COSMO GM 2013
Eidgenössisches Departement des Innern EDIBundesamt für Meteorologie und Klimatologie MeteoSchweiz
13
a) b)
c) d)
Figure 6: The location of the encounter is shown as intersection of the red lines (location of vertical cross sections) in a) and as cross (+) in b) and c). Figure 6.a) shows the topography with overlayed wind vectors for wind at flight altitude. The red lines show the location of cross sections alongwind and crosswind. b) shows ths synoptic ground chart of DWD (n.d.), c) horizontal wind speed and wind vectors at flight altitude and d) the vertical cross section of W, TH and 2pvu at 15 UTC.
Location of Cross Sections & Wind Vectors @ Flightaltitude 21.01.2011 10 UTC
−15 −10 −5 0 5
−8
−6
−4
−2
0
2
4
6
8
10
250
500
750
1000
1250
1500
1750
2000
2250
2500
2750
3000
Wind speed [m/s], wind vectors, 2pvu @ 10910m (flight altitude) 21.01.2011 10 UTC
−15 −10 −5 0 5
−8
−6
−4
−2
0
2
4
6
8
10
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
met
er
Vertical Cross Section of vertical Wind W, Theta and 2pvu 21.01.2011 15 UTC
2
2
22
2
2
2
2
2
2
2
2
22
22
2
22
2
2
52
275275
280280
280
285285
285
290290
290
295295
295
300300
300
305305
305
310310
310
315315315
320320
320
325325
325 330330
330
335335
335340
340
340345
345
345350
350
350 355355
355 360360360
365365
365
370
370370
370
375
375
−6.5 −6 −5.5 −5 −4.5 −40
2000
4000
6000
8000
10000
12000
−2.1−1.95−1.8−1.65−1.5−1.35−1.2−1.05−0.9−0.75−0.6−0.45−0.3−0.1500.150.30.450.60.750.91.051.21.351.51.651.81.952.1
Samstag, 31. August 2013