Gridded Localized Aviation MOS Program (LAMP) Guidance for Aviation Forecasting

Judy E. Ghirardelli and Bob Glahn

National Service Meteorological Development Laboratory

August 2, 2011

15th Conference on Aviation, Range, and Aerospace Los Angeles, CA

1 LAMP/Gridded LAMP (GLMP) • LAMP: . Guidance of sensible weather at stations . Guidance of (≥ 1 CTG strike) on grid • GLMP: Gridded observations and Gridded LAMP forecasts of: . . Dewpoint . Ceiling Height (100’s of ft) . (miles) . Other elements later • Status: . GLMP running experimentally at NCEP as of 9/28/2010, will be operational 9/27/2011 . Data available in Experimental NDGD . Available in MDL’s test Web Coverage Service . Exp. GLMP grids can be brought into AWIPS via the LDM data feed . Can be used in the creation of NWS digital aviation products . Images available on LAMP web page . Will be available via the NextGen 4-D Data Cube 2 Gridded LAMP Details • Gridded LAMP analyses of observations – for checkout and verification . Temperature and Dewpoint: • Observations from METAR, , synoptic stations, C-MAN, tide gauges, and moored buoys (Roughly 10,000 – 12,000 observations per hour) • Error estimates of temperature and dewpoint available in gridded format . Ceiling Height and Visibility: • Observations from METAR • Gridded Analysis of LAMP forecasts . Temperature and dewpoint: • Continuous values are analyzed • < 1500 LAMP stations (input points) in CONUS, so SREF forecasts are used to augment LAMP forecasts . Ceiling Height and Visibility: • Categorical values are converted to continuous values • No augmentation

3 Technique: MDL Gridding Technique used in Gridded MOS, with modifications Gridded LAMP: 0-25 hours Temperature Dewpoint

Ceiling Height Visibility

4 Temperature and Dewpoint Gridded Verification

• Compared GLMP vs. GMOS . 0600 UTC GLMP vs 0000 UTC GMOS . 1800 UTC GLMP vs 1200 UTC GMOS (in paper) • Data Sample: November-December 2010 • Area: CONUS, 2.5-km grid • Variables: Temperature (Dewpoint results shown in paper) • Verification using two methods: . GLMP 0-hr gridded observations . RTMA • Computed: . MAE over all gridpoints . Fractional Improvement in MAE of GLMP over GMOS at each gridpoint • Purpose: to determine if GLMP improves on GMOS as LAMP improves on MOS

5 MAE over all gridpoints 0600 UTC Gridded LAMP Verification Temperature - All Regions 4.50

4.00

3.50

3.00

2.50

2.00 0600 UTC GLMP vs. GLMP Gridded Obs 1.50

Mean Absolute Absolute Error (F) Mean 1.00 0000 UTC GMOS vs. GLMP Gridded Obs 0.50

0.00 09 12 15 18 21 00 03 06 Valid Time (UTC) 6 MAE over all gridpoints 0600 UTC Gridded LAMP Verification Temperature - All Regions 4.50

4.00

3.50

3.00

Error (F) 2.50

2.00 0600 UTC GLMP vs. GLMP Gridded Obs 1.50 0000 UTC GMOS vs. GLMP Gridded Obs

Mean Absolute Absolute Mean 1.00 0600 UTC GLMP vs. RTMA Gridded Obs 0000 UTC GMOS vs. RTMA Gridded Obs 0.50

0.00 09 12 15 18 21 00 03 06 Valid Time (UTC) 7 Fractional improvement in MAE of GLMP over GMOS at gridpoints (3-hr forecast from 0600 UTC GLMP)

GLMP Temperature GLMP Temperature Verified w/ 0-hr GLMP Verified w/ RTMA

8 Ceiling Height and Visibility Station Verification

• No Gridded ceiling/visibility verifying observations other than from Gridded LAMP  verification done at stations only • Two cycle times: 0600 (1800 UTC in paper) • Verified for Nov-Dec, 2010 • Verifying at stations . At LAMP 300 stations • Purpose: to see if GLMP inter- polated to the stations is as good as actual LAMP at the stations. . At 115 non-LAMP stations  115 stations where LAMP station forecasts were not available, but obs were available  Mimics with-held data testing

9 Threat Score 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 1 3 5 7 9 8 7 6 5 4 3 2 1 0 Threat Height Scores0600UTC Ceiling < (AtLAMPand NON-LAMP stations)

01 21 41 61 81 02 22 425 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 Projection (Hour)Projection

PERSIST (115 non-LAMP stns)(115 non-LAMP PERSIST stns) non-LAMP (115 GLMP stns) (300 LAMP PERSIST stns) LAMP (300 GLMP stns) LAMP (300 LAMP 1000 FT

Threat Score 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 1 3 5 7 9 8 7 6 5 4 3 2 1 0 Threat Scores (AtLAMPand NON-LAMP stations)

0600 01 21 41 61 81 02 22 425 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 Projection (Hour) Projection UTC VisibilityUTC < 3 MILES

PERSIST (115 non-LAMP stns)(115 non-LAMP PERSIST stns) non-LAMP (115 GLMP stns) (300 LAMP PERSIST stns) LAMP (300 GLMP stns) LAMP (300 LAMP

Future Improvements  Persistence Effect for temperature and dewpoint: . Uses spatial detail seen in 0-hr GLMP observations (10,000-12,000 input points) as well as additional MOS input points to provide more spatial detail in GLMP forecast analyses  Redeveloped ceiling height  Extension to 30 hrs  Adding Stations . 119 new MOS stns (116 in CONUS) . 306 Marine stns . 274 Canadian stns . 4 new TAF stns . 15 military stns (13 in CONUS)

12 Threat Score 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 1 2 3 4 5 6 0600 UTC GLMP Verification 7 Ceiling Height < 1000 Feet (11 Canadian Stations) 8

9 10 Projection (Hour) Projection 11 12 13

14 15

16

17

18 19 LPNEW GLMP PERSISTENCE CURRENT GLMP 20 21 22 23 24 25 Threat Score 0.7 0.8 0.9 0.1 0.2 0.3 0.4 0.5 0.6 1 0 2 4 6 8 9 8 7 6 5 4 3 2 1 0600 UTC GLMP Verification (11 Canadian Stations)

Visibility< 3 miles 01 21 41 61 81 02 22 425 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 Projection (Hour) Projection

LPNEW GLMP PERSISTENCE CURRENT GLMP Summary  GLMP running experimentally at NCEP for temperature, dewpoint, ceiling height, and visibility; GLMP useful in the creation of digital aviation products  GLMP for temperature and dewpoint : . Overall, GLMP shows improvement over GMOS, as expected, independent of verifying data (GLMP gridded obs or RTMA) . By gridpoint, GMOS temperature is better than GLMP temperature in Western Region for some gridpoints, some projections • Potential solution developed and soon to be tested. Would be implemented in 2012.  GLMP for ceiling height and visibility: . Overall, GLMP interpolated to LAMP stations shows no degradation compared to LAMP guidance at stations, as expected . GLMP interpolated to non-LAMP stations potentially worse than GLMP interpolated to LAMP stations, as expected . Improvements were seen in Canada after adding Canadian stations (to be implemented in future)  LAMP Website: . http://www.nws.noaa.gov/mdl/gfslamp/gfslamp.shtml  Contact: . [email protected]

15