ECMWF: towards coupled earth system modelling and assimilation - - PowerPoint PPT Presentation

▶
ecmwf towards coupled earth system modelling and
SMART_READER_LITE
LIVE PREVIEW

ECMWF: towards coupled earth system modelling and assimilation - - PowerPoint PPT Presentation

Jul 18, 2023 231 likes •553 views

ECMWF: towards coupled earth system modelling and assimilation Including Copernicus services Colleagues and Peter Bechtold ECMWF, Shinfield Park, RG2 9AX, Reading, UK EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS October 29, 2014 1

slide-1
SLIDE 1

October 29, 2014 EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

1

ECMWF: towards coupled earth system modelling and assimilation

Including Copernicus services Colleagues and Peter Bechtold

ECMWF, Shinfield Park, RG2 9AX, Reading, UK

slide-2
SLIDE 2

October 29, 2014

Challenges

  • Predicting European weather a month ahead

(“beyond ten days”)

  • Understanding the influence of the tropics on

global weather

  • Understanding the influence of the stratosphere
  • n the troposphere during winter
  • Describing the interactions between

atmosphere, oceans, sea-ice, land surface and composition

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

200 mb winds on 15 March 2014

Day 6 Day 5 Day 4 Day 2 Day 3

EU summer heatwave 2-4 week ahead

2

slide-3
SLIDE 3

October 29, 2014

ECMWF 2016-2025 strategy: overview

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

3

  • Focus on high-impact weather, regime transitions and global-scale anomalies
  • Integrated ensemble at high resolution at 5km by 2025

Performant Earth-System model and analysis which includes relevant processes (variables/observables) with an estimate of process/model uncertainty

  • Scalable computation
  • Environmental information services: Copernicus
slide-4
SLIDE 4

October 29, 2014

IFS components (as of Oct 2016)

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

4

HRES 4DVAR 9km EDA-25 4DVAR 18km HRES FC 9km (d0-10) ENS-51 FC 18km (d0-15) SEAS4-51 FC 80km atmosphere land waves atmosphere land waves currents sea-ice ORAS4-25 3DVAR-FGAT 1 degree currents sea-ice atmosphere land waves

Initial conditions Forecasts Model components Model components

CERA reanalysis

slide-5
SLIDE 5

Improvement in cloud cover skill – the last decade

Evolution of skill of the HRES forecast at day 5, expressed as relative skill compared to ERA-Interim (12 month running mean) TCC See also Haiden et al (2015) ECMWF Newsletter 143

36r4 38r2 41r1 36r1 38r1 37r2 31r1 32r3 37r3 40r1

slide-6
SLIDE 6

ATMS and Metop-B MHS added in clear skies

6 EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

Observation changes: the rise of all-sky!

All-sky assimilation of humidity sounding channels on SSMIS All-sky assimilation of all four MHS (transferred from clear-sky) GMI and AMSR-2 added in all-sky F18, all-sky over snow, MWHS-2

  • Growing importance of microwave

humidity observations (MHS, ATMS, MWHS- 2, SSMIS, AMSR2, GMI, SAPHIR).

  • Extending this to infrared water

vapour information.

  • Revisiting all-sky microwave

temperature observations.

  • Also investigating radar, lidar, and

possible lightning observations (EarthCARE, Aeolus, GOES-R, MTG).

slide-7
SLIDE 7

October 29, 2014

New observations

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

slide-8
SLIDE 8

October 29, 2014

Grid res. HRES ENS LegA LegB/M’ly 4DV inner loops 1st 2nd 3rd EDA Outer 1st 2nd 128 km 64 km 32 km 16 km 9 km TL639 TCo639 TL319 TCo319 TL1279 TCo1279 TL255 TL399 TL255 TL319 TL399 TCo639 TL159 TL191 TL159 TL191 TL255

41r2: Resolution upgrade – 8 March 2016

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

41r1  41r2

8

(D10 D15) workshop/

slide-9
SLIDE 9
  • Two grids use same

spectral truncation

  • Spectral fields, here the
  • rogrophay look nearly

the same, but not quite (more detail!!) why?

From Tl1279 (16 km) to TCo1279 (9 km)

slide-10
SLIDE 10

October 29, 2014

Resolution upgrade: cubic grids->octahedral reduced Gaussian grid

2N+1 gridpoints to N waves : TL linear grid 4N+1 gridpoints to N waves : T

c cubic grid

  • Mathematically more correct in the presence of cubic

non-linearities in the equations

  • Less numerical filtering – almost no numerical

diffusion, no dealiasing

  • Better mass conservation
  • Less expensive than the equivalent linear grid

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

slide-11
SLIDE 11

N640 TL1279 (linear, ~16km) ≈ 2.1 million grid points per level N1280 TC1279 (cubic, ~8km) ≈ 8.5 million grid points per level O1280 TCo1279 (cubic, ~9km) ≈ 6.6 million grid points per level (octahedral cubic reduced Gauss. grid)

N24 O24

Resolution upgrade: octahedral reduced Gaussian grid

slide-12
SLIDE 12

OPER CUBIC China China

Improvements: ….

Strong reduction of spurious grid-scale rainfall events (LSP)

Frequency

  • f rain

events >20mm/6h

slide-13
SLIDE 13

Oper 41r1, 3h precip High Res 41r2, 3h precip Satellite IR image

Tropical Cyclone Soudelor Aug 2015

  • Instability in Numerics due to departure point calculation

in the semi-Lagrangian advection, leading to unrealistic tropical cyclone structures

Improvements: resolve instabilities in Numerics (Advection)

slide-14
SLIDE 14

October 29, 2014

Ocean surface currents at various resolutions

14 EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

Eddy resolving Eddy permitting Eddy parameterising

43r1, ENS

slide-15
SLIDE 15

October 29, 2014

Thermal coupling of ocean

Coupled ocean-atmosphere simulations are exposed to the problem

  • f initial shock as the Atmosphere and the rest of earth surface is

not yet in balance with the ocean. Data assimilation 4D-VAR uses OSTIA SST

  • 1/20 degree SST from OSTIA
  • 1/4 degree SST from NEMO

– dynamic, ocean tendencies – uncertainties

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

OSTIA SST 1/20 degree is applied for 4- days and then relaxed to 0 gradually taper- down from day 4 to day 8 From day 8 onwards FULL COUPLING The PARTIAL COUPLING works well only in the short range as ocean eddies are assumed stationary FULL COUPLING uses the dynamic ocean to advect eddies from day 0

NEMO SST, FULL-COUP. OSTIA SST OSTIA SST + NEMO ΔSST

Day0 Day4 Day8

ORCA025 1/4 degree 1m with 8-layer in top 10m OSTIA SST 1/20 degree OSTIA 1/20 deg (5km) SST field has details of the eddies not resolved by ocean models

slide-16
SLIDE 16

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

October 29, 2014

Coupled ocean vs uncoupled simulation

16

4-day forecast SSTs from the coupled forecast initialised at 0UTC on 6 July 2014 at the location of a buoy with approximate position 22°N, 128°E. uncoupled coupled (Rodwell et al, ECMWF Technical Report 759, 2015) Tropical cyclone Neoguri with TCo1279 (HRES) Buoy observation at 22°N, 128°E

slide-17
SLIDE 17

October 29, 2014

(Balsamo et al, ECMWF Newsletter 137, Tellus-A, 2012)

Lake model 2015, one of the new Earth System components

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

5 10 15 20 25 6/1/15 7/1/15 8/1/15 IFS41r1_lake_T IFS41r1_lake_T_old OSTIA-OSI-SAF_lake_T

Lake Baikal

17

LAKE COVER FRACTION

Forecast of 2m temperature are improved in proximity of lakes and coastal areas Why also coastal areas, these are not Lakes ?!...... cause before if land- sea mask>0.5 then only land point

slide-18
SLIDE 18

October 29, 2014

A new and simple lightning parametrisation - also for use in data assimilation

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

18

  • P. Lopez 2016 MWR
slide-19
SLIDE 19

Climatological AOD at 550 nm distribution CAMS vs operational climatology (based on Tegen et al. 1997)

19

  • Aerosol climatology computed using the CAMS-Interim reanalysis (Flemming et al. 2016)
  • Some highlights:

– Larger Sea Salt radiative forcing (~1 W/m2 more reflection at TOA over oceans) – Changes in biomass burning seasonal cycle (up to 20 W/m2 difference in total SW absorption locally) – Changes in dust distribution, higher on Sahara and Taklamakan, lower on Indian Ocean and Australia – Anthropogenic emissions lower over Europe, higher over E Asia

slide-20
SLIDE 20

October 29, 2014

And real time Analysis and Fc of Aerosols within Copernicus Atmospheric monitoring system recent fires in Chile

20 EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

slide-21
SLIDE 21

October 29, 2014

Shallow cumulus Deep cumulus Stratocumulus dry BL

zi Zi=Zcb

Mixed layer M moist (conv) M dry (turb) K K K K K K

Evaluating forecasts against observations

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

One of the flights during CSET

OPER NEW CSET, the Cloud System Evolution in the Trades – July/August 2015 (University of Washington and Miami) NARVAL (Next-generation Aircraft Remote sensing for Validation Studies) – MPI-M (Dec 2013/Jan 2014) RH

dry cloudy

Towards a more unified turbulence, convection, cloud interaction

slide-22
SLIDE 22

October 29, 2014

Ensemble and stochastic physics: Pattern generator

22 EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

slide-23
SLIDE 23

October 29, 2014

Ensemble and stochastic physics: Perturbed parameter distributions

23 EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

slide-24
SLIDE 24

October 29, 2014

The Scalability Challenge

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

Observations Models Volume 20 million = 2 x 107 5 million grid points 100 levels 10 prognostic variables = 5 x 109 Type 98% from 60 different satellite instruments physical parameters of atmosphere, waves, ocean Observations Models Volume 200 million = 2 x 108 500 million grid points 200 levels 100 prognostic variables = 1 x 1013 Type 98% from 80 different satellite instruments physical and chemical parameters of atmosphere, waves, ocean, ice, vegetation

Today: Tomorrow:

→ Factor 10 per day → Factor 2000 per time step

(10-day forecast today = 1440 time steps, but more time steps with increased resolution)

24

slide-25
SLIDE 25

October 29, 2014

Simple compute projection (only resolution)

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

Ensemble Single

≈ M€ electricity/year Power limit (Bauer et al. 2015, Nature)

25

slide-26
SLIDE 26

October 29, 2014

Copernicus climate services: Reanalysis provides a truly global view…

  • ERA-Interim estimates for

2014 are slightly cooler than those from station data alone

  • Mainly due to the Antarctic
  • Consistent with independent
  • bservations of sea-ice extent
  • ERA-5 now underway
slide-27
SLIDE 27

October 29, 2014

This leaves us with uncertainty in the uncertain times …….

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

27

  • Dynamical core: spectral or Finite Volume Method
  • Physics: which additional prognostic equations (in microphysics-convection, Ozone and dust+sea salt

aerosols coupled with radiation?

  • Mult-layer snow scheme
  • First wish of our satellite microwave assimilation people is to have “prognostic convective snow”
  • Data assimilation: which hybrid method, continue with ensemble 4DVar (maintaining TL/AD model

veeery work intensive but still pays off)

  • Scalability
slide-28
SLIDE 28

October 29, 2014

“Dry” evaluation of the FVM vs IFS

EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

28

  • C. Kuehnlein and P. Smolarkiewicz
slide-29
SLIDE 29

October 29, 2014

TCo1279 total column liquid water (12h simulation at 9 km)

29

(hydrostatic, with deep convection parametrization, 450s time-step, 240 Broadwell nodes, ~0.75s per timestep)

TCo1279 orography

slide-30
SLIDE 30

October 29, 2014

TCo7999 total column liquid water (12h simulation at 1.3 km)

30 EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS

(hydrostatic, no deep convection parametrization, 120s time-step, 960 Broadwell nodes, ~10s per timestep)

TCo7999 orography

The latest spectral transform model news …