Ralph Petersen 1 and Robert M Aune 2

NearCasting Severe Convectionusing Objective Techniques that

Optimize the Impact of Sequences of GOES Moisture Products

Ralph Petersen1and Robert M Aune2

1 Cooperative Institute for Meteorological Satellite Studies (CIMSS), University of Wisconsin – Madison, Madison, Wisconsin, [email protected]

2 NOAA/NESDIS/ORA, Advanced Satellite Products Team, Madison, Wisconsin, [email protected]

Improving the utility of GOES productsin operational forecasting

What are we trying to improve?Short-range forecasts of timing and locations of severe thunderstorms

- especially hard-to-forecast, isolated summer-time convection

What are we trying to correct?- Poor precipitation forecast accuracy in short-range NWP

- Lack of satellite moisture data over land in NWP

- Time delay in getting NWP guidance to forecasters

-Dominance of parameterizations over observations

in increasingly complex short-range NWP

- Excessive smoothing of mesoscale moisture patterns

in NWP data assimilation- Smoothing of upper-level dynamics and dryness

- Lack of objective tools for use by forecasters

in detecting pre-convective environments 1-6 hours in the future

Basic premises - NearCasting Models Should:Update/Enhance NWP guidance:

Be Fast (valid 0-6 hrs in advance)Be run frequently Can avoid ‘computational stability’ issues of longer-range NWP

Use all available data quickly:“Draw closely” to good data Avoid ‘analysis smoothing’ issues of longer-range NWP

Be used to anticipate rapidly developing weather events:“Perishable” guidance products – need rapid delivery Avoid ‘computational resources’ issues of longer-range NWP

Run Locally? – Few resources needed beyond comms, users easily trained

Focus on the “pre-storm environment”Focus on the “pre-storm environment” - - Increase Lead Time / Probability Of Detection (POD)Increase Lead Time / Probability Of Detection (POD) and and

Reduce False Alarm Rate (FAR)Reduce False Alarm Rate (FAR)

Goals: - Increase the length of time that forecasters can make good use of quality observations which can supplement NWP guidance for their short range forecasts

- Provide objective tools to help them do this

0 1 5 6 hours

Fill the Gap

To increase usefulness, GOES Sounder products are available to forecasters as imagesProducts currently available include: - Total column Precipitable Water (TPW) - 3-layers Precipitable Water (PW) - Stability Indices, . . .

DPI Strengths and Current Limitations

+ Image Displays speed comprehension of information in GOES soundings, and

+ Data Improve upon Model First Guess,

but - - -

- Products used only as observations, and - Currently have no predictive component - Data not used in current NWP models

A Specific Objective: Extend the benefits of GOES Sounder Derived Product Images (DPI)

Moisture Data to forecasters

GOES 900-700 hPa PW - 20 July 2005

TPW NWP Guess Error Reduction using GOES-12

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

BIAS(mm) SD(mm)

Vertical Layer

% R

edu

ctio

n w

ith

GO

ES

-12

900-700 hPa

700-300 hPa

Lower-tropospheric GOES Sounder Derived Product Imagery (DPI)

3 layers of Precipitable Water

Sfc-900 hPa900-700 hPa700-300 hPa

0700 UTC

GOES 900-700 hPa Precipitable Water - 20 July 2005

1000 UTC

1300 UTC

1800 UTC

after initial storms have developed, cirrus blow-off can mask lower-

level moisture maximum in subsequent DPI satellite products

Small cumulus developing in boundary layer later in day can

also mask retrievals

Forecasters need tools

that:

Preserve high-

resolution data

and

Show the future state

of the atmosphere

And,

- Data used (and combined) directly - no ‘analysis smoothing’ – Retains observed maxima and minima and extreme gradients

- Spatial resolution adjusts to available data density

- GOES products can be projected forward at full resolution - even as they move into areas where clouds form and subsequent IR observations are no longer available

- Data can be tracked (aged) through multiple observation times

- New data are added at time observed – not binned / averaged

- Use best aspects of all available data setse.g., Cloud Drift winds can be combined with surface obs cloud heights to create a ‘partly cloudy’ parcel with good height and good motion

e.g., Wind Profiler data can be projected forward and include accelerations

What are the benefits of a Lagrangian NearCasting approach?

Extending a proven diagnostic approach to prediction

- Quick (10-25 minute time steps) and minimal resources needed

- Can be used ‘stand-alone’ or to ‘update’ other NWP guidance

DATA DRIVEN

How it works:

Instead of interpolating randomly spaced moistureobservations to a fixed grid (and smooth data) and then using gridded wind data to determine changes of to calculating moisture changes at the fixed grid points, the Lagrangian approach interpolated wind data to each of the moisture obs

Dry

Moist

13 April 2006 – 2100 UTC900-700 hPa GOES PW

0 Hour NearCast

Lagrangian NearCast

How it works:

Instead of interpolating randomly spaced moistureobservations to a fixed grid (and smoothes data) and then using gridded wind data to determine changes of to calculating moisture changes at the fixed grid points, the Lagrangian approach interpolated wind data to each of the moisture obs and then moves the 10 km data to new locations, using dynamically changing wind forecasts with ‘long’ (10-15 minute) time steps.


0 Hour Ob Locations

Lagrangian NearCast


1 Hour NearCast Obs

How it works:


Lagrangian NearCast


2 Hour NearCast Obs

How it works:


Lagrangian NearCast


3 Hour NearCast Obs

How it works:


Lagrangian NearCast

13 April 2006 – 2100 UTC900-700 hPa GOES PW3 Hour NearCast Image

Dry

Moist

10 km data, 10 minute time steps

How it works:

Instead of interpolating randomly spaced moistureobservations to a fixed grid (and smoothes data) and then using gridded wind data to determine changes of to calculating moisture changes at the fixed grid points, the Lagrangian approach interpolated wind data to each of the moisture obs and then moves the 10 km data to new locations, using dynamically changing wind forecasts with ‘long’ (10 minute) time steps.

The moved moisture ‘obs’ values are then

periodically transferred back to an ‘image’ grid,

but only for display.

Lagrangian NearCast

Obser

vatio

n: A pr

eferre

d loc

ation

for r

apid

deve

lopmen

t of c

onve

ction

is u

nder a

nd alon

g

edge

s of m

id-lev

el dr

y ban

ds as

they

mov

e

over

loca

lized

low-le

vel m

oistu

re m

axim

a

4 3 2 1 0

Elevation (km

)

-20 -10 0 10 20 30Temperature (C)

If parcels are lifted from either the top or bottom of the inversion layer, they both become cooler than surroundings and sink

Determining if the atmosphere is conducive to convective storms

STABLE Vertical Temperature Structure showswarm air over colder air

WHY?

Before we go farther,

A Quick Tutorial

When the entire layer islifted with moist bottom and top dry, the inversionbottom cools less than top and it becomes absolutely unstable – producing auto-convection

4 3 2 1 0

Elevation (km

)

-20 -10 0 10 20 30Temperature (C)

Moist

Dry


But if sufficient moisture is added to bottom of an otherwise stable layer and the entire layer is lifted, it canbecome Absolutely Unstable.

Latent heat

added

4 3 2 1 0

Elevation (km

)

-20 -10 0 10 20 30Temperature (C)

Moist

Dry

Lifted layer can become Absolutely Unstable.This will cause thelayer to overturnand the moist air

will raise very rapidly


Requires sequence of low-level moistening,

mid/upper-level drying and descent (needed

to trap lower-level moisture and prevent

convection from occurring too quickly)

- followed by decreased

convergence aloft.

So, we really need to monitor not only the increase of low level moisture,

but ,more importantly, to identify areas where Low-level Moistening and Upper-level Drying will occur

simultaneously

When the entire layer islifted with moist bottom and top dry, the inversionbottom cools less than top and it becomes absolutely unstable – producing auto-convection

A case study was conducted for the hail storm event of 13 April 2006that caused major property damage across southern WisconsinThe important questions are both where and when severe convection will occur, and

Where convection will NOT occur?

NWP models placed precipitation too far south and west - in area

of large-scale moisture convergence

10 cm Hail24 Hr Precip

A case study was conducted for the hail storm event of 13 April 2006that caused major property damage across southern WisconsinThe important questions are both where and when severe convection will occur, and

Where convection will NOT occur?


0 Hour NearCast


0 Hour NearCast

Initial Conditions

0 Hour Moisture for 2100UTC

From 13 April 2006 2100UTC

GOES Derived Product Images

900-700 hPa PW700-300 PW 2115 UTC Sat Imagery

InitialGOES DPIs

Valid2100 UTC


0 Hour NearCast


0 Hour NearCast

2115 UTC Sat Imagery

NAM 03h Forecast of 700-300 hPa and 900-700 hPa PW valid 2100 UTC

Initial Conditions




900-700 hPa PW700-300 PW

InitialGOES DPIs

Valid2100 UTC

Comparison with NWP conditions


0 Hour NearCast


0 Hour NearCast


NearCasts

0 Hour Projection for 2100UTC



900-700 hPa PW700-300 PW

Low-LevelMoistening

Low-LevelMoistening

0 hr LagrangianNearCasts

ofGOES DPIs

Valid2100 UTC


2 Hour NearCast


2 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Low-LevelMoistening


ofGOES DPIs

Valid2300 UTC


4 Hour NearCast


4 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Low-LevelMoistening


ofGOES DPIs

Valid0100 UTC


6 Hour NearCast


6 Hour NearCast


NearCasts

6Hour Projection for 0300UTC



900-700 hPa PW700-300 PW

Low-LevelMoistening


ofGOES DPIs

Valid0300 UTC


0 Hour NearCast


0 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying

Mid-LevelDrying


ofGOES DPIs

Valid2100 UTC


2 Hour NearCast


2 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying


ofGOES DPIs

Valid2300 UTC


4 Hour NearCast


4 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying


ofGOES DPIs

Valid0100 UTC


6 Hour NearCast


6 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying


ofGOES DPIs

Valid0300 UTC


0 Hour NearCast


0 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying

Mid-LevelDrying


ofGOES DPIs

Valid2100 UTC

Validation of rapid moisture changein NE Iowa using GPS TPW data

GOES Observed 900-300 PW = 18+6 = 24GPS TPW = 25

It this

real ?

Validation of Local NE Iowa

moisture maximumwith GPS TPW

data


2 Hour NearCast


2 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying


ofGOES DPIs

Valid2300 UTC


GOES Observed 900-300 PW = 20+5 = 25GPS TPW= 28

It this

real ?



data


4 Hour NearCast


4 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying


ofGOES DPIs

Valid0100 UTC


GOES Observed 900-300 PW = 22+4 = 26GPS TPW= 30

It this

real ?



data


6 Hour NearCast


6 Hour NearCast


NearCasts




900-700 hPa PW700-300 PW

Mid-LevelDrying


ofGOES DPIs

Valid0300 UTC


GOES NearCast 900-300 PW = 10 + 3 = 13GPS TPW = 16 13

It this

real ?



data


0 Hour NearCast


0 Hour NearCast

Vertical Moisture Gradient(900-700 hPa GOES PW - 700-500 hPa GOES PW)

0 Hour NearCast for 2100UTCFrom 13 April 2006 2100UTC



ofGOES DPIs

Valid2100 UTC

Differential Moisture Transport Creates Vertical

Moisture Gradients

Formationof

ConvectiveInstability

Differential Moisture Transport Creates

Convective Instability


2 Hour NearCast


2 Hour NearCast





ofGOES DPIs

Valid2100 UTC

Formationof



Moisture Gradients


4 Hour NearCast


4 Hour NearCast





ofGOES DPIs

Valid2100 UTC

Formationof



Moisture Gradients


6 Hour NearCast


6 Hour NearCast






ofGOES DPIs

Valid2100 UTC

Formationof



Moisture Gradients


0 Hour NearCast


0 Hour NearCast

Initial Conditions





InitialGOES DPIs

Valid2100 UTC

Examples like this show that:

- The GOES DPI moisture products provide good data

- The Lagrangian NearCasts produce useful and frequently updates about the FUTURE stability of the atmosphere

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

A number of additional modifications have been made to the image products more useful in operations:

• Eliminate displays in areas without NearCast ‘data’• “Look” more like satellite observations

• Add uncertainty into longer NearCast images by increasing smoothing through NearCast period

Also:

•Want to show the impact of frequent data updates available only from geostationary systems

Areal influence of each predicted trajectory point expandswith NearCast length (Longer-range NearCasts have more uncertainly and should therefore show less detail)

Run Time > 18 UTC 19 UTC 20 UTC 21 UTC Valid

18UTC

19UTC

20 UTC

21 UTC

22 UTC

23 UTC

00 UTC

01 UTC

02 UTC

03 UTC

Convectively Stable

Convectively Unstable

19ZNewData

No Data

20ZNewData

21ZNewData

Derived Vertical Moisture DifferenceGray areas indicate regionswithout trajectory data fromlast 6 hours of NearCasts

Verifying Radar

RUC 6 hr Cloud Top ForecastValid 03 UTC

Run Time > 20 UTC 21 UTC 22 UTC 23 UTC Valid

20 UTC

21 UTC

22 UTC

23 UTC

00 UTC

01 UTC

02 UTC

03 UTC

04 UTC

05 UTC

Convectively Stable

21ZNewData

No Data

22ZNewData

23ZNewData

Derived Vertical Moisture Difference

Convectively Unstable

Verifying Radar

Repeated insertion of hourly GOES data improves and revalidates NearCasts

- Data can be inserted (combined) directly retaining resolution and extremes - NearCasts retain useful maxima and minima – image cycling preserves old data Forecast Images agree with storm formations and provide accurate/timely guidance

Summary – An Objective Lagrangian NearCasting Model - Quick and minimal resources needed

- Can be used ‘stand-alone’ or to ‘update’ other NWP guidance

DATA DRIVEN at the MESOSCALE

Goals met: Provides objective tool to increase the length of time that forecasters

can make good useof detailed GOES moisture data in

their short range forecasts(augment smoother NWP output)

Expands value of GOES moisture products from observations

to short-range forecasting tools

Testing planned in NWS offices

Stable

Unstable

Dry

Moist

Differential Moisture Transport Creates

Convective Instability


0 Hour NearCast


0 Hour NearCast

Initial Conditions





InitialGOES DPIs

Valid2100 UTC

Examples like this show that:

- The GOES DPI moisture products provide good data

- The Lagrangian NearCasts produce useful and frequently updates about the FUTURE stability of the atmosphere

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Next steps:

24/7 runs

Need WFO evaluation:

Is this a useful forecasting tool? ( If so, when and when not? )What additional tuning is needed?

What other parameters should be included (LI, CAPE, . . .)

How should it be tested? ( Web-based initially and/or directly to AWIPS? )

Let’s talk more

Comparison of GOES (left) and AIRS (right) data coverage around 0700 UTC 20 July 2006. Times and lateral limits of AIRS overpasses shown.

For GOES: Details of moisture maximum (warm colors) which was initially over Iowa and subsequently moved eastward to support convection over WI and IL are clearly identified in spatially continuous data.

For POES: No AIRS data were available over IA (indicated by white areas), due to combination of: 1) cloud obscurations (e.g., over MN and western IA in later 0900 UTC data), and 2) data gaps between successive orbital paths (e.g., central and eastern IA).

Note: Neither radiosonde nor aircraft moisture data would have been available around 0700 UTC for this area.

Why we need GOES (not just POES) data for detecting small-scale convection, An unforecasted mesoscale Derecho which initially moved from MN across south-central WI, decayed and then re-intensified south of Chicago

Ralph Petersen 1 and Robert M Aune 2

Documents

Transcript of Ralph Petersen 1 and Robert M Aune 2