Seamless Probabilistic Forecasts for Civil Protection: from week to minutes

Similar documents
Nowcasting systems and user applications; experiences from INCA, PROFORCE and ASIST

INCA CE: Integrating Nowcasting with crisis management and risk prevention in a transnational framework

INCA-CE: The Challenge of Severe Weather Warnings. Yong Wang, ZAMG

LAM-EPS activities in RC LACE

LAM-EPS activities in LACE

Ingo Meirold-Mautner Yong Wang Alexander Kann Benedikt Bica Georg Pistotnik

WMO/WWRP FDP: INCA CE

LAM-EPS activities in RC LACE

Application and verification of ECMWF products 2012

INCA-CE achievements and status

The Hungarian Meteorological Service has made

Integrating Nowcastingwith crisis management and risk prevention in a transnational framework (INCA-CE)

AROME Nowcasting - tool based on a convective scale operational system

Winter Storm Briefing #3 Monday Evening Wednesday Evening

Xinhua Liu National Meteorological Center (NMC) of China Meteorological Administration (CMA)

Developments towards multi-model based forecast product generation

The benefits and developments in ensemble wind forecasting

Application and verification of ECMWF products 2009

Road weather forecasts and MDSS in Slovakia

JMA Contribution to SWFDDP in RAV. (Submitted by Yuki Honda and Masayuki Kyouda, Japan Meteorological Agency) Summary and purpose of document

Recent progress in convective scale Arome NWP system and on-going research activities

Probabilistic Weather Prediction

Interaction of boundary-layer turbulence and convection in the AROME model: Preliminary results

Strategic Radar Enhancement Project (SREP) Forecast Demonstration Project (FDP) The future is here and now

South African Weather Service. Description of Public Weather and Warning Services. Tshepho Ngobeni. 18 November 2013

Seamless nowcasting. Open issues

Application and verification of ECMWF products 2015

Preliminary results. Leonardo Calvetti, Rafael Toshio, Flávio Deppe and Cesar Beneti. Technological Institute SIMEPAR, Curitiba, Paraná, Brazil

ALADIN forecast experience in Croatia

Application and verification of ECMWF products 2017

Application and verification of the ECMWF products Report 2007

Severe weather warnings at the Hungarian Meteorological Service: Developments and progress

The convection-permitting COSMO-DE-EPS and PEPS at DWD

TC/PR/RB Lecture 3 - Simulation of Random Model Errors

Stochastic methods for representing atmospheric model uncertainties in ECMWF's IFS model

Perspectives on EPS at EUMETNET and at AEMET

ADL110B ADL120 ADL130 ADL140 How to use radar and strike images. Version

Mesoscale meteorological models. Claire L. Vincent, Caroline Draxl and Joakim R. Nielsen

recent Météo-France work on AROME-EPS Robert Osinski & François Bouttier - CNRM, Toulouse

Aurora Bell*, Alan Seed, Ross Bunn, Bureau of Meteorology, Melbourne, Australia

Current Issues and Challenges in Ensemble Forecasting

Nowcasting for the London Olympics 2012 Brian Golding, Susan Ballard, Nigel Roberts & Ken Mylne Met Office, UK. Crown copyright Met Office

Application and verification of ECMWF products 2015

Application and verification of ECMWF products 2017

OBJECTIVE CALIBRATED WIND SPEED AND CROSSWIND PROBABILISTIC FORECASTS FOR THE HONG KONG INTERNATIONAL AIRPORT

Application and verification of ECMWF products in Norway 2008

Probabilistic Winter Weather Nowcasting supporting Total Airport Management

Low-end derecho of 19 August 2017

Nesting and LBCs, Predictability and EPS

At the start of the talk will be a trivia question. Be prepared to write your answer.

1. INTRODUCTION * Figure 1. National Weather Service Storm Prediction Center (SPC) storm reports for December 1, 2006.

Heavy Rainfall Event of June 2013

MSG FOR NOWCASTING - EXPERIENCES OVER SOUTHERN AFRICA

Utilising Radar and Satellite Based Nowcasting Tools for Aviation Purposes in South Africa. Erik Becker

Application of ALADIN/AROME at the Hungarian Meteorological Service

How ECMWF has addressed requests from the data users

FLORA: FLood estimation and forecast in complex Orographic areas for Risk mitigation in the Alpine space

Application and verification of ECMWF products 2016

AROME-EPS development in Météo-France

FLOOD/SCS EVENT, APRIL 28 MAY 4

WG4: interpretation and applications

On the use of radar rainfall estimates and nowcasts in an operational heavy rainfall warning service

Model Output Statistics (MOS)

Denver International Airport MDSS Demonstration Verification Report for the Season

LAMEPS activities at the Hungarian Meteorological Service. Hungarian Meteorological Service

0-6 hour Weather Forecast Guidance at The Weather Company. Steven Honey, Joseph Koval, Cathryn Meyer, Peter Neilley The Weather Company

Weather Briefing. Coastal Storm Monday-Wednesday Dec 8-10, National Weather Service. Prepared 12/8/14 6:30 AM

High Resolution Numerical Weather Prediction for High Impact and Extreme Weather Events in 2014 across Southern California

The ECMWF Extended range forecasts

FLASH FLOOD FORECASTING AND EARLY WARNING IN SLOVENIA. Sašo Petan Hydrological Forecast Department, Slovenian Environment Agency (ARSO)

Deutscher Wetterdienst

C. Gebhardt, S. Theis, R. Kohlhepp, E. Machulskaya, M. Buchhold. developers of KENDA, ICON-EPS, ICON-EDA, COSMO-D2, verification

State of the art of wind forecasting and planned improvements for NWP Helmut Frank (DWD), Malte Mülller (met.no), Clive Wilson (UKMO)

Behind the Climate Prediction Center s Extended and Long Range Outlooks Mike Halpert, Deputy Director Climate Prediction Center / NCEP

Judit Kerényi. OMSZ-Hungarian Meteorological Service P.O.Box 38, H-1525, Budapest Hungary Abstract

Winter Weather. National Weather Service Buffalo, NY

Austria (ZAMG) Implementation and development of AROME. Upgrade ALADIN-LAEF. Research on land surface data assimilation, e.g. ASCAT.

Diabatic processes and the structure of the warm conveyor belt

Assimilation of radar reflectivity

Flood Forecasting. Fredrik Wetterhall European Centre for Medium-Range Weather Forecasts

Severe Weather Potential for Southeast Texas

ECMWF products to represent, quantify and communicate forecast uncertainty

István Ihász, Máté Mile and Zoltán Üveges Hungarian Meteorological Service, Budapest, Hungary

Catalysing Innovation in Weather Science - the role of observations and NWP in the World Weather Research Programme

DETECTION AND FORECASTING - THE CZECH EXPERIENCE

Convection Nowcasting Products Available at the Army Test and Evaluation Command (ATEC) Ranges

Thunderstorm-Scale EnKF Analyses Verified with Dual-Polarization, Dual-Doppler Radar Data

U.S. WIND, SCS, FLOOD, WINTER WEATHER

The analysis and nowcasting system INCA. T. Haiden, A. Kann, K. Stadlbacher, G. Pistotnik, M. Steinheimer, F. Wimmer, C. Wittmann

Exploring stochastic model uncertainty representations

Weather forecasts and warnings: Support for Impact based decision making

Montréal, 7 to 18 July 2014

Convective-scale NWP for Singapore

Application and verification of ECMWF products 2011

Severe Weather Watches, Advisories & Warnings

Probabilistic weather hazard forecast guidance for transoceanic flights based on merged global ensemble forecasts

Dijana Klarić RC LACE Program Manager

Flood Risk Forecasts for England and Wales: Production and Communication

IWT Scenario 2 Integrated Warning Team Workshop National Weather Service Albany, NY October 31, 2014

Description of the case study

Transcription:

Seamless Probabilistic Forecasts for Civil Protection: from week to minutes Yong Wang, Clemens Wastl, Andre Simon, Mihaly Szűcs ZAMG and HMS

An EU project Bridging of Probabilistic Forecasts and Civil Protection Project: 2013 2015; Website: www.proforce.eu Project aim Weather service and civil protection agency work together on: An innovative seamless probabilistic forecasting system in time (week to hour) and space (30km to 1km) tailored to civil protection

Seamless forecasts & actions ECMWF-EPS ~ 30 km resolution 15 days Civil Protection RESPONSE Potential severe weather event ALADIN-LAEF 11km resolution 3 days Civil Protection PREPARE More precise forecast AROME-EPS 2.5 km resolution 1,5 days Civil Protection GO-action Detailed forecast Ensemble INCA 1 km resolution 0-6 hours

Ensemble prediction systems ECMWF EPS 30km AROME EPS, AROME RUC and EN-INCA AROME EPS 2,5km EN-INCA 1 km AROME RUC 1,2km ALADIN-LAEF 11km

ALADIN-LAEF

Surface perturbation

Atmospheric perturbation

Stochastic surface physics

AROME-EPS AROME-EPS ensemble size 16 + 1 Δx / vertical levels 2.5 km / 90 coupling runs per day IC perturbation model perturbation ECMWF EPS 2 / 4 runs (+30h forecast) EDA + EnJk +seda SPPT / psppt / ssppt: total tendencies (SPPT) micro-physics shallow convection turbulence

AROME-EPS: stochastic physics SPPT_global SPPT_adapt Influence of tapering function 70000 60000 50000 40000 30000 20000 10000 0-1 -0.9-0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Total tendencies Radiation scheme AROME-EPS Partial tendencies Radiation scheme δtt 1 δtt Shallow convection scheme δtt 2 δtt, δrr 2 δtt, Turbulence scheme δuu 2, VV 2 δtt δtt 3 δtt, δrr 3 δtt, δuu 3, VV 3 δtt Microphysics scheme δtt 1 δtt (1 + PP) Shallow convection scheme ( δtt 2 δtt, δrr 2 δtt, δuu 2, VV 2 ) (1 + PP) δtt Turbulence scheme ( δtt 3 δtt, δrr 3 δtt, δuu 3, VV 3 ) (1 + PP) δtt Microphysics scheme 4 dddd dddd = δttii δtt ii=1 δtt 4 δtt, δrr 4 δtt (1 + PP) dddd dddd = ( δtt 4 δtt, δrr 4 δtt 4 dddd dddd = δttii δtt ii=1 ) (1 + PP) dddd dddd =

Ensemble INCA: Details EN-INCA Precipitation Analysis Nowcasting blending NWP Measurement error * Representation error * Radar error * Interpolation error * Extrapolation error Convective effects* NWP 05.08.2016 errors / Folie 12 uncertainties and/or Multi-Modelapproach * Measurement error* Representation error * Interpolation error * Ensemble background fields Error in NWP-trend -> LAEF Stochastical blending with LAEF NWP errors / uncertainties Perturbed META (Objectively Optimized Forecasting System) Analyse Nowcasting blending NWP Temperatur, wind

Estimation of radar error Error stdev

Ensmeble INCA From Error motion vectors to a nowcasting ensemble Error motion vector based on the error of previous motion vectors (e.g. last 30-45 minutes) random variation of error motion vector (within Gaussian distr.) yields spread beginning at the very first forecast time step

Examples of ensemble precipitation analysis

Statistical calibration

Webpage: www.echo-proforce.eu

Webportal

Threat by impact! Threat level: Impact definded by Civil Protection! - Combination of probability and impact - 3 warning levels: yellow, orange, red - Also the forecast time plays a role: The farther into the future the more severe a potential event has to be to reach threat level - General decay of forecast quality with increasing lead time is acknowledged in that way

Case studies Storm Niklas March 30 April 2 2015 10 days before the event: ECMWF-EPS First signals of a heavy storm event are already visible - first awareness actions in the forecasting centres. ECMWF forecast for Tuesday March 31, 18 UTC (Init. 2015032212 UTC). Probability of gusts > 60km/h (up) and > 80km/h (right). ECMWF Meteogram for the capital city of St. Pölten (Init. 2015032212 UTC). Some members exceed the 100km/h threshold for this location.

Case studies Storm Niklas March 30 April 2 2015 3 days before the event: LAEF LAEF with a horizontal resolution of 11km includes more orographic details and confirms the forecast of ECMWF-EPS. CP receives first official warnings warning level was upgraded to orange. Local authorities are informed, first preparation actions are launched. Official warning of ZAMG issued on March 27 12 UTC. Warning level was upgraded to orange (annuality < 3 times a year). LAEF forecast for Tuesday March 31, 18 UTC (Init. 2015032812 UTC). Probability of gusts > 60km/h (upper), > 80km/h (middle) and > 100km/h (lower).

Case studies Storm Niklas March 30 April 2 2015 1 day before the event: AROME-EPS Convection permitting EPS system AROME with 2.5km resolution and 48 hours forecasting range. Warning level orange was extended in time, basic warning was increased to 100km/h for the western parts of Lower Austria - final preparedness and prevention actions in CP are launched. AROME-EPS forecast for Tuesday March 31, 17 UTC (Init. 2015033015 UTC). Probability of gusts > 60km/h (left), > 80km/h (middle) and Ensemble max. wind speed (right).

Case studies Storm Niklas March 30 April 2 2015 Nowcasting: Ensemble-INCA 1km resolution, 12h forecasting range, 15min update interval.warning level orange was extended in time, basic warning was increased to 100km/h for the western parts of Lower Austria - final preparedness and prevention actions in CP are launched. Ensemble-INCA forecast for Tuesday March 31, 14 UTC (Init. 2015033112 UTC). Probability of gusts > 80km/h (left) and Ensemble max. wind speed (right).

Case studies Storm Niklas March 30 April 2 2015 Feedback from Civil Protection - Very good wind forecasts of the seamless system - First awareness already several days in advance (much earlier than deterministic) - Additional information about uncertainty helped to focus the CP forces - Very high correspondence of warnings and number of operations Map of Upper Austria (left) and Lower Austria (middle) with marked operations due to strong wind. A comparison with the seamless forecast of the event (probability of exceeding 60km/h) shows the very high correspondence between operations and forecast.

A severe windstorm over Hungary 6 days before the event (Tuesday, 28 April 2015) Maximum gust (can exceed 90 km/h) ECMWF-EPS: Somogy county 1. Windstorm from NW 2. In the morning (peak at 09 UTC), 3. North of the county (Lake Balaton) 4. probability: 20-30% over 70km/h

A severe windstorm over Hungary 4 days before the event (Tuesday, 28 April 2015) Maximum gust (70-80 km/h) ECMWF-EPS: Change! In the evening (peak at 18 UTC)! probability: 50-60% over 70km/h! Intensity weaker!

A severe windstorm over Hungary 3 days before the event (Tuesday, 28 April 2015) Maximum gust (70-80 km/h) ECMWF-EPS: Probability: 40% over 70km/h! No probability for 90km/h. Intensity even weaker! Weak thunderstorm! EPS Maximum for thunderstorm index Average intensity thunderstorms Showers, weak thunderstorms

A severe windstorm over Hungary 1 days before the event (Tuesday, 28 April 2015) Maximum gust (70 km/h) ALADIN-EPS: Change! not realistic Somogy county peak at 21 UTC! probability: 50% over 70km/h, but no > 90 km/h Intensity weaker! Shower, thunderstorm possible

A severe windstorm over Hungary The day of the event, in the morning (Tuesday, 28 April 2015) Maximum gust (70 km/h) ALADIN-EPS: No important change! The maximum gust even a bit weaker. Thunderstorm possible, weak to normal! EPS Maximum for thunderstorm index

The event 1400 UTC (Tuesday, 28 April 2015) INCA-EPS 3h nowcasts: Frontal precipitation approaching the area of Balaton, convective pre-frontal cells coming from south INCA EPS Wind analysis and nowcast: The wind will reach gale intensity at Balaton within 3h Analysis 3h forecast

Caution! Day of the event 1700 UTC (Tuesday, 28 April 2015) Wind analysis and nowcast: The wind already turned at Balaton and intensified (70-90 km/h gusts) as planned. But the INCA nowcasts indicate that the wind can reach 100 km/h as well. This was not forecast by ECMWF and ALADIN-EPS!!! 100 km/h Analysis +3h Forecast

The event 1700 UTC (Tuesday, 28 April 2015) Observations confirmed! Windstorm at Balaton is already strong, the front reached the inner part of the County as well. The stable cold air inhibits convection. Heavy cells are only at the south. Siófok at Balaton: gale Kaposvár: after rain, windy roll-clouds waves

The event 2000 UTC (Tuesday, 28 April 2015) INCA-EPS 3h precipitation nowcasts: Heavy precipitation is aready out of the Somogy County. But the gale at Balaton further intensified!!! But INCA can sometimes overestimate the gust speed. We might verify it... EPS INCA maximum precipitation INCA wind analysis gusts > 110 km/h

The event 2000 UTC (Tuesday, 28 April 2015) Wind gusts and average windspeed at Balatonőszöd (central basin and southern shore of the Lake) 139 km/h 2000 UTC

IMPACT, after the event (Wednesday, 29 April 2015) Impact of : Damaged power lines on the railway at Balaton, buildings, ships in the harbour of Siófok. About 17 rescue actions required.

Summary and Follow up In the frame of PROFORCE, 4 weather and civil protection partners from Austria and Hungary have worked together on: An innovative seamless probabilistic forecasting system in time and space tailored to civil protection, for: improvements in the accuracy and timeliness of severe weather warnings improved preparedness and decision making procedures in civil protection agencies cooperation between weather and civil protection authorities transnational cooperation on warning Adaptation of the developed system for further use in other European countries and in other application area