Indirect Use of NWP in Nowcasting

Zentralanstalt für Meteorologie und Geodynamik

Indirect Use of NWP in Nowcasting

Yong Wang, ZAMG, Austria

With contribution from Bica, Meyer, Kann, Pistotnik, Xie etc.


21.04.23(Präsentation)

Folie 221.04.23(Präsentation)

Folie 2Nowcasting systems use NWP indirectly

(Pierce et al., 2004)




(Wilson et al., 2010)




Model Name Organization Country Spatial Resolution

Temporal Resolution Available

Times of Day Run (UTC)

Length of Forecst (hours)

General Description

ABOMLAM1km Environment Canada Canada 1 Km 15 min Every 15 min Max 6 h

Adaptive Blending of Observation and Models

using GEM LAM1k

ABOMREG Environment Canada Canada 15 km 15 min Every 15 min Max 6 h

Adaptive Blending of Observation and Models

using GEM Regional

INTW Environment Canada Canada 1 and 15

km 15 min Every 15 min Max 6 hINTegrated Weighted

Model using LAM1k, GEM Regional and Observations

INCA ZAMG Austria 1 km 1 hour Every hour 18 hours

Downscaled ECMWF forecasts as a first guess and applies corrections according to the latest

observation.

(George Isaac, 2011)



Folie 5Integrated Nowcasting through Comprehensive Analysis

Analyses and Nowcasting

Surface observations

Satellite obsercations

Geoinformation data

NWP forecasts

INCAData QC, Integration,

optimisation

Radar observations

Radiosonde

(INCA reference see Haiden et al., 2011)



Folie 6INCA-CE ─ A Central European Nowcasting Initiative

• EU funded Nowcasting project• 16 partners from 8 European countries

• Hydro-Met services• Research institutions• Public authorities

• Project budget: 4.7 million US$• Project duration: Apr 2010 – Sep 2013

• ZAMG leading• www.inca-ce.eu

• Application orineted nowcasting R&D, rapid INCA, user oriented nowcast product/grafics

• Nowcasting application in crisis managment and risk prevention in civil protection, operational Hydrology and road management

• Nowcasting based transnational warning strategy



Folie 7

Domain size600 x 350 km

Elevation range100 - 4000 m

ResolutionHorizontal: 1 kmVertical: 150 mTime: 15 min – 1h

Update frequency5 min – 1h

Availability+ 20 min … +30 min

INCA configuation and topography


INCA uses NWP products 21.04.23(Präsentation)

Folie 8

There is limited interdependency between the fields. In the nowcasting of temperature the cloudiness analysis and nowcast are taken into account. The surface cooling caused by convective cells due to the evaporation of precipitation enters the analysis and nowcasting of temperature.

Derived fields include convective parameters such as the lifted condensation level (LCL), or CAPE. Snowfall line and ground temperature are computed for nowcasts of precipitation type (snow, rain, snow–rainmix, freezing rain).

(Haiden et al., 2011)



Folie 9

Observation analysis

Blending

Nowcast, including advection, initiation, growth and decay of convection

Nowcast products

Ensemble Nowcasting

Comparison: INCA (NWP based) – VERA (non-NWP)

Indirect use of NWP in Nowcasting in:


Observation analysis 21.04.23(Präsentation)

Folie 10

Short range NWP forecasts are usually used as first guess in the observation analysis in nowcasting



Folie 11

The analysis of temperature starts with an NWP short-range forecast as a first guess, which is then corrected based on observation–forecast differences.

Observation analysis in INCA: Temperature

Corrections to the first guess are computed based on the differences ΔTk between the observed and NWP temperatures at station locations.

Similar to Temperature , NWP forecasts are used as first guess in humidity and wind analysis.



Blending 21.04.23(Präsentation)

Folie 12

)()()()()( ,,,, NWPjjNWPnowcastiinowcastblend

The blended forecast is calculated as the weighted sum of the extrapolation and NWP. The forecast values are combined using a time-varying weighting function which is derived from the measured performances.

To choose an appropriate quality measure is crucial. The weighting method can be linear, exponential, or the introduction of stochastic noise.



Folie 13Overview of blending

(Atencia and Germann, 2010)



Folie 14Overview of blending



Folie 15Blending in B08FDP

(B08FDP/RDP report, 2009)



Folie 16Blending in INCA

To obtain a continuous sequence of forecast fields, a transition from the extrapolation forecast to the NWP forecast is constructed through a prescribed weighting function that gives full weight to the extrapolation forecast during the first 2 h and decreases linearly to zero at 6 h.

Attempts to improve upon the fixed weighting by making the time scale of the transition dependent on the magnitudes of NWP and nowcasting errors has as yet not shown any benefit.

Update frequency:ECMWF 12 h (available at +9 h)ALARO5 6 h (available at +5 h)Nowcasting 5,15 min (available at +20…25

min)(Haiden et al., 2011)



Folie 17Blending in INCA



Folie 18Nowcast in INCA: convection

„INCA convective Nowcasting“:

For each „convective girdpoint“ (i.e., with CAPE > 50 J/kg in a certain area):

• Initiation?

• Growth?

• Decay?

KTT

kgJCIN

VIS

sMOCON

kgJCAPE

trig 2

200

5.0

/102

1006

KTT

kgJCIN

hmmRR

sMOCON

kgJCAPE

trig 2

200

2.0

/102

506

hmmRR

sMOCON

kgJCAPECAPE

0.0

/0

100,min0

All the index are computed

from NWP products.

(Pistotnik et al., 2011)



Folie 19Nowcast in INCA: verification

RMSE of convective Nowcast with ALADIN vs. RMSE of translation-Nowcast (all Termine, t0+3h)

Green: improvement by convective nowcast




RMSE of convective Nowcast with AROME vs. RMSE of translation-Nowcast (all Termine, t0+3h)

Green: improvement by convective nowcast






Folie 22Nowcast in INCA: temperature and humidity

In the case of temperature and humidity, Lagrangian persistence explains only a small part of the total temporal variation, and variations due to the diurnal cycle become dominant.

The temperature nowcast is based on the trend given by the NWP model and computed for each grid point from a recursive relationship.

TINCA(t0) temperature at the analysis time

Thus, the INCA temperature nowcast is the latest analyzed temperature plus the temperature change predicted by the NWP model, multiplied by fT.

This factor is parameterized as a function of the cloudiness forecast error of the NWP model.

If the NWP model underestimates the cloudiness compared to the INCA cloudiness analysis and nowcast, it will tend to overpredict temperature changes, and vice versa.




Folie 23Nowacst products

Precip typeSnowfallSnow/Rain mixRainFreezing rain

Lightning rate



Folie 24Nowcast product

Many nowcast products are diagnosed using nowcating forecasts in conjunction with NWP products, which provide the estimate ofatmospheric structure:

• Visibility: liquid water content, aerosol content

• Lightning rate: updraught velocity in convective clouds

• Precipiatation type: snowfall line, 3D T and Q, cloud information

• Icing potential: T and wind

(Golding, 1998; Haiden et al., 2011)



Folie 25Ensemble Nowcasting based on det. NWP

Short Term Ensemble Prediction System- NWP blend

Decompose NWP into a cascade Decompose the rainfall field into a cascade Use radar field to estimate stochastic model parameters Calculate the skill of the NWP at each level in the cascade using the

correlation between NWP and radar Blend each level in the radar & NWP cascades using weights that are a

function of the forecast error at that scale and lead time For each forecast

Add noise component to the deterministic blend, the weight of the noise is calculated using the skill of the blended forecast

Combine the cascade levels to form a forecast

(Seed, 2011)

Details in presentation of Peter Steinle



Folie 26Ensemble Nowcasting based on NWP EPS

(Kober et al., 2010)


En-INCA: INCA + ALADIN-LAEF 21.04.23(Präsentation)

Folie 27

Experimental:

En-INCA = blending (INCA, ALADIN-det., LAEF) = blending ( prob. convective nowcast, AROME, LAEF)

En-INCA = INCA as control + downscaled spread (LAEF)



Folie 28

Ensemble Size 16 +1

horizontal resolution

18 km

Vertical resolution

37 levels

Runs/day 2 (00,12UTC)

Forecast range 60h

Time step 720s

Coupling-model ECMWF EDA/SV EPS

Coupling-update

6h

LAEF: Limited Area Ensemble Forecasting

Atmosphere perturbation: Blending ALADIN Bred + ECMWF EDA/SV

Surface perturbation: Non-Cycling surface Perturbation

Model perturbation: multi-physics

ALADIN-LAEF



Folie 29Comparison: INCA and VERA analysis

There are wo Nowcasting systems in Vienna:

VERA (Vienna Enhanced Resolution Analysis, Steinacker et al. 2006) is NWP independent and based on variational principle applied to higher-order spatial derivatives. It uses a fingerprint technique to integrate conceptual / climatological information, or upscaled radar data.

INCA relies on NWP model products and remote sensing data to interpolate between observations.


INCA vs. VERA 21.04.23(Präsentation)

Folie 30



Folie 31INCA vs. VERA

Weather dependent!


Conclusions 21.04.23(Präsentation)

Folie 32

NWP is widely used in Nowcasting systems indiectly:

• Observation analysis and nowcast products• Blending• Nowcast including advection, initiation, growth and decay of convection• Ensemble Nowcasting

Progress in NWP in the last years, e.g. advanced data assimilation technique, comprehensive model physics and cloud resolving model; assimilation of very dense observations in time and space, like radar, GPS etc., there will be more and more use of NWP directly and indirectly in Nowcasting.



Folie 33

Thanks!

Documents

Indirect Use of NWP in Nowcasting