Revisiting the freezing spray forecast algorithms used in Canada Une revisite des algorithmes de prévision d embruns verglaçants utilisés au Canada

Transcription

1 Revisiting the freezing spray forecast algorithms used in Canada Une revisite des algorithmes de prévision d embruns verglaçants utilisés au Canada par Serge Desjardins Mark Pilon Ted Mcildoon* National Lab for Marine and Coastal Meteorology (aka MAC Lab) *Atlantic Storm Prediction Centre Halifax, Nova Scotia CMOS 2012 Montréal, Québec

2 FREEZING SPRAY PROJECT - PART 1 : Taking the responsibility of the freezing spray program - a) Generates the YQB_FRZSP maps (Golfe, Hudson) Modified Stallabrass - b) Duplication of the YQB_FRZSP on various domains (NAtl, Artic) - PART 2 : Collaboration with USA (Seattle-Marine Workshop) : Matt Glazewski, Aimee Fish - Adding the Alaska domain - Adding the Overland algorithm (used by USA) - Verifying various Stallabrass algorithm code - i) Original Stallabrass (from Ross Brown s code) - ii) Modified Stallabrass (from Ross Brown s code) = YQB_Modified_Stallabrass - iii) Modified Stallabrass (from SCRIBE [interface&matrices] = mixture of i) & ii) = not pure - PART 3 : Operimental script to generate 1PAN & 4PAN charts on various domain and with different flavors - Overland - Original, Fetch Factor (wave,wind) - Stallabrass - Modified (sig wave, wind wave), original, YQB, SCRIBE, rel. Vessel speed = 5 m/soriginal - PART 4 : Subjective Evaluation - Collaboration USA & Canada - With Various SPCs (discussion on this) Page 2

3 Domains Activity by June 2012 Products de-activated Alaska Kane? Until June East- Arctic Products activated in coming summer months Hudson Labrador? in July? Until August? Until August Beaufort Baffin Gulf of St-Lawrence N. Atlantic? in July Page 3

4 ATMO/WAVE INPUTS to compute Freezing Spray rate Domain ATMO WAVE Comments Dom Res. Dom Res. Alaska GLB 35 km PAC 0.5 deg Arctic GLB 35 km ARC 0.5 deg < 0.5 deg Wave Hudson REG 15km ARC 0.5 deg Prob deg Wv N. Atl. REG 15km NAT 0.55 deg 0.15 deg Gulf Coupled-REG 15km GSL 0.05 deg Page 4 Ice timelydepend.

5 Freezing Spray Overland Algorithm SAWADA (from COMAR) FREEZING SPRAY TABLE WIND SPEED (KTS) PPR = Icing Predictor (m o Cs -1 ) V a = Wind Speed (m s -1 ) T f = Freezing point of seawater (usually -1.7 o C or -1.8 o C) T a = Air Temperature ( o C) T w = Sea Temperature ( o C) Modified NIL NIL F F F F F F F SF T - 5 NIL F F F F F F F SF SF E - 6 F F F F F F F SF SF SF M - 7 F F F F F F F SF SF SF P - 8 F F F F F F SF SF SF SF E - 9 F F F F F SF SF SF SF SF R - 10 F F F F F SF SF SF SF SF A - 11 F F F F SF SF SF SF SF SF T - 12 F F F F SF SF SF SF SF SF U - 13 F F F F SF SF SF SF SF SF R - 14 F F F F SF SF SF SF SF SF E - 15 F F F SF SF SF SF SF SF SF F = COMAR CODE FOR MODERATE FREEZING SPRAY SF = COMAR CODE FOR SEVERE FREEZING SPRAY Page 5 Sawada condition

6 Overland Algorithm (calm down please!) Over most of Canadian Waters (Golfe, Hudson, Arctic) PPR = V a ( ) ( ) ( ) - = = F, T f T a V a T a V a T a Fetch Factor or % Spray Flux availability FETCH _ PPR = Ł Avai Max SprayFlux * PPR = SprayFlux ł Wave Sig ( UV ) 13 Wind * 2 PPR Page 6

7 Freezing Spray Stallabrass Algorithm q Considers air temperature, water temperature, wind speed, and wave height q q Assume 20 knots for relative to the wave vessel speed (SCRIBE) Predicts the average icing rate o 1200 Events from 138 different vessels. Data collected by Jim Stallabrass 1968 to 1980 o Each event described by 67 fields o Icing categorized by severity: Trace, Lights, Moderate, Heavy Page 7

8 Summary of the Algorithms FREEZING SPRAY TABLE Category Extreme Severe Moderate Cloud of possibility generated by the other variables WIND SPEED (KTS) NIL NIL F F F F F F F SF T - 5 NIL F F F F F F F SF SF E - 6 F F F F F F F SF SF SF M - 7 F F F F F F F SF SF SF P - 8 F F F F F F SF SF SF SF E - 9 F F F F F SF SF SF SF SF R - 10 F F F F F SF SF SF SF SF A - 11 F F F F SF SF SF SF SF SF T - 12 F F F F SF SF SF SF SF SF U - 13 F F F F SF SF SF SF SF SF R - 14 F F F F SF SF SF SF SF SF E - 15 F F F SF SF SF SF SF SF SF Light F = COMAR CODE FOR MODERATE FREEZING SPRAY SF = COMAR CODE FOR SEVERE FREEZING SPRAY SAWADA Wind Speed SAWADA : F (Wind Speed, Temperature) OVERLAND : F (Wind Speed, Temperature, Sea Surface Temperature) STALLABRASS : F (Wind Speed, Temperature, Sea Surface Temperature, Wave, RH) Page 8

9 Why Numerical Model Outputs are still a forecast guidance Cold frontal passage Jan 18/19, 2012 Page 9

10 Model Output Temperatures Matrix Points Page 10 St. Paul Island OBS

11 January 19/00Z (24HR) to 20/00z (48HR) OVERLAND F(T,UV,TM) STALLABRASS F(T,UV,TM,WV,HR) SAWADA F(T,UV) Page 11

12 3:00 am Fcst Jan. 18/2012 CABOT STRAIT. GALE WARNING IN EFFECT. WIND SOUTHEAST 15 TO 20 KNOTS INCREASING TO 30 LATE THIS MORNING WITH GUSTS TO 50 IN THE WRECKHOUSE AREA. WIND BECOMING SOUTHWEST 35 LATE THIS AFTERNOON THEN DIMINISHING TO NORTHWEST 25 THURSDAY MORNING. WIND DIMINISHING TO NORTHWEST 15 THURSDAY EVENING. SNOW CHANGING TO RAIN EARLY THIS MORNING AND ENDING LATE THIS AFTERNOON. FLURRIES BEGINNING NEAR MIDNIGHT. FOG PATCHES FORMING EARLY THIS MORNING AND DISSIPATING EARLY THIS EVENING. TEMPERATURES MINUS 1 RISING TO PLUS 5 THIS AFTERNOON THEN FALLING TO MINUS 3 THURSDAY AFTERNOON. Page 12

13 3:00 am Fcst Jan. 19/2012 CABOT STRAIT. GALE WARNING IN EFFECT. FREEZING SPRAY WARNING IN EFFECT. WIND WEST 35 TO 45 KNOTS DIMINISHING TO NORTHWEST 30 THIS MORNING AND TO 15 EARLY THIS EVENING. WIND VEERING TO NORTHEAST 15 NEAR MIDNIGHT THEN DIMINISHING TO LIGHT FRIDAY MORNING. WIND BECOMING SOUTHEAST 15 FRIDAY EVENING. FLURRIES ENDING NEAR NOON. FREEZING SPRAY ENDING THIS MORNING. TEMPERATURES NEAR MINUS 3. Page 13

14 Conclusions : Freezing Spray - findings Graphics are very helpful Better than nomogram (point by point) Gives forecaster the whole picture Provides a variety of algorithms Use/evaluation hampered by lack of reports Based on biased model temperatures Page 14

15 Conclusions : Next Steps v More Validation work needs to be done v Big Problem : Lack of observations v Developing a global product to get more ship obs? v Canada/USA Collaboration continue v Probably a joint Operational Validation next year v An evolving file with METAREA.. To be continued v The SPRAY GENERATION (spray which could eventually freeze) comes from the ship s movement. v How can we take this into account in our guidance products? v Stallabrass has it but Page 15