Ideas for Improving Hurricane Forecasts

MiamiJune 20, 2008

Alexander E.MacDonald

OAR DAA

Director Earth System Research Laboratory

Talk Summary

1. Observing systems: UAS

2. Understanding phenomena: Air-Sea interaction.

3. Modeling: FIM, NIM

There are a great variety of UAS.

Aerosonde Global Observer Zephyr

Global Hawk

AEM in situ

Altair Predator B

NOAA UAS Program Funded!

• Program will apply to all NOAA missions

• This presentation: Tropical Storms

• Program Manager: Marty Ralph, FY 2008

Robbie Hood, FY 2009

• Budget: $ 3.3 M in 2008, $ 6.3 M in 2009

Examples of Hurricane Missions for UAS:

1.Global Hawk to loiter above the hurricane.

2.Aerosonde into the hurricane at low levels.

3.Zephyr over hurricanes for days at a time.

4.WISDOM (Small super-pressure balloons)

NOAA is working with NASA with the goal of using the Global Hawk as a platform for hurricane recon.

Global Hawk could loiter above hurricanes for extended periods, and:

• Dropsondes

• Radar sampling

• Scatterometry

• Radiometry

Aerosonde was launched from NASA’s Wallops Flight Facility at 14:08 EST on Friday November 2, 2007.

Mission endurance was 17 hours 27 minutes and resulted in 7.5 hours of data collection in the core of the hurricane at altitudes less than 100m.

NOAA/NASA Noel Aerosonde UAS mission

AerosondeP-3 (3500 m)

Take-off from Wallops Flight Facility (KWAL) 1600 UTC

Rendezvous with P-3 at storm center 0525 UTC

KMCF

“Eye” loitering with vertical soundings

(90-1500 m)

Nominal altitude for the Aerosondeprior to entering the storm’s core circulation was between 300-600m.

Near the center, UAS altitude was ~150m.

ZEPHYR - Long Duration (Many days) UAS

• Built by QinetiQ Corporation (UK)

• Problem: Payload only 4 kg !!!!!!!

Zephyr HALE UAS • Long endurance from solar-electric power

– Weeks/months flight endurance– Operating altitude 50,000 - 65,000ft– No environmental impact

• Drop micro-sondes into hurricanes or monitor oceanic weather events for long periods

• Satcom capable – commercial Iridium• Low cost platform for payloads, plus:

– No regular maintenance– No operating base needed– Launch and recover from ship’s deck– Fewer personnel required– No fuel distribution network needed

• Zephyr flies for longer for less money– ~$1m production platform cost

Ultra-light Dropsonde

* ESRL/GSD issued RFP for dropsonde in 2006

* Spec: Closest to 20 gram sonde wins

* Winner: Applied Research Associates

* Delivery of test system: October 2008

Dropsonde Mechanical Design• Foam housing• 3.5 inch diameter• 2.3 grams• Air pathways for sensors

• Populated housing with attached streamer

• 38.3 grams

Dropsonde Electronics

• Packaging – Open cell foam w/ streamer• Microprocessor – NXP LPC2138• Transmitter – TI CC1100, 300MHz – 1GHz• Pressure – Intersema MS5534B• Humidity – Sensiron SHT75• Temperature – Platinum RTD• GPS – Ublox LEA-5S• Power – ER14250, Li 1/2AA 3.6VDC

Production System Concepts• Streamer-based dropsonde dispenser

Zephyr – in-situ measurement using drop-sondes

Lightweight drop-sondes dispensed into

developing weather event or storm at

intervals to measure PTU and wind speed

Hurricane WISDOM: 2008 Test

Briefing to NWS

Silver Spring, May 6, 2008

Alexander E.MacDonald, Justyna Nicinska

NOAA Earth System Research Laboratory

Small balloons called “tetroons” can be inexpensive and easy to handle.

There are commercially available devices that can report GPS based 3 D location to a satellite in real time.

Some weigh less than an ounce.

Up to 1000 balloons could be launched to eliminate initial state uncertainty.

Flood the data poor region with balloons.

2008 Test Plan

National Weather Service Interactions

(1) Data made available to NCEP (EMC, NHC) in real time via internet.

(2) Possible need for access to helium at RAOB sites.

Goal: Release balloons between August 17 and August 22 to improve track.

Aug 22

Aug 18

Aug 19Aug 20

Aug 21

Aug 17

(1) Aug 17 model ensemble identifies target zones for Aug 22.

(2) Balloons are released from Cape Verde, Florida, Cancun.

2008 Test Plan

• Test funded at $ 1 M by DHS.• Test period: August 15 to October 1.• Balloons: 100 at 700 mb, 100 at 200 mb.• Cape Verde and Barbados: 40 balloons each.• Deployment teams: 6 teams of 2 people.• Team travel kits: Duffel bag with 20 balloons, plus valves.• Team travel: Commercial air.• Operations center: ESRL.• Nominal target storms: 2

Talk Summary

1. Observing systems: UAS

2. Understanding phenomena: Air-Sea interaction.

3. Modeling: FIM, NIM

Sea Spray ProfilingEstimates Based on Fairall-Banner Model

Scanhead in UAS aircraft• Radiometer scanhead in Altair

Test with Katrina (2005)KATRINA 2005082700 HWRF Maximum Wind Speed (m/s) from track info

0

10

20

30

40

50

60

70

80

90

0 10 20 30 40 50 60 70

Forecast Hour

Win

d S

pee

d (

m/s

)

Best track estimate

Control

Spray ft=1 ss=1.0

Spray ft=1 ss=10.0

Spray ft=1 ss=1.0 w/ momentum

Talk Summary

1. Observing systems: UAS

2. Understanding phenomena: Air-Sea interaction.

3. Modeling: FIM, NIM

An early look at the forecast performance of the Flow-Following

Icosahedral Model (FIM)

Presented at ESRL Theme Presentation on “Global Weather Assimilation and Modeling”, W 7 May 2008

Forecast skill must be adequate to establish FIM’s readiness to contribute to Global Ensemble Forecast System

SOME SPECIFICS we use to verify

FIM running regularly twice daily at ESRL - Initial conditions from 0000 and 1200 UTC initialized fields for the NCEP Global Forecast System (GFS) operational runs - Computational polygons are ~ 30km in diameter, covering the globe - 50 computational layers (bottom: earth’s surface GFS terrain height; top: 20 hPa) - GFS Physics Important Point: A single case proves nothing!

http://fim.noaa.gov > Global or CONUS

Example: Tropical Depression 03W Western Pacific (FIM and GFS forecasts

from 1200 UTC Tue 6 May 2008)

Joint Typhoon Warning Center Warning#2 1200 UTC 7 May 2008

00h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Tue 7 May 2008

+

24h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Wed 7 May 2008

48h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Thu 8 May 2008

72h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Fri 9 May 2008

96h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Sat 10 May 2008

120h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Sun 11 May 2008

144h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Mon 12 May 2008

168h forecast of sea-level pressure (hPa - 1000)FIMGFS

Valid time: 1200 UTC Tue 13 May 2008

Road Map of the high-end Earth System Modeling

2006-2013: The “Geodesic Grid” Planet Simulator, 3rd generation non-hydrostatic finite-volume model with 1-4 km or finer resolution.

Primary model for:• Medium Range Weather Prediction 0 to 2 weeks, including

improve hurricane track and intensity.

• Seasonal to Inter-annual Prediction 2 weeks to 2 years

• Decadal to Centennial projection

A project between GFDL, ESRL, AOML and NCEP to develop the leading global model in the world.

Alexander.E.MacDonald@noaa.gov sandymacd@comcast.net