Implementation of IFS cycle 46R1

#IFS46r1 #newfcsystem

Description of upgrade

IFS Cycle 46r1 is an upgrade with many scientific contributions, including changes in data assimilation (both in the EDA and the 4DVAR), in the use of observations, and in modelling. The new cycle only includes meteorological changes; there are no technical changes, e.g. new resolutions.

The page will be updated as required. It was last changed on 17 May 2019 

For a record of changes made to this page please refer to  Document versions.

Further information and advice regarding the upgrade can be obtained from User Support.

News

17 May 2019 

• The IFS cycle 46r1 scorecards are now available.
• Cycle 46r1 test data are available in ecCharts and as ENS meteograms.
• Recording and slides of second cycle 46r1 webinar are available.
• The implementation of the new cycle is confirmed for Tuesday 11 Jun 2019. The first operational run using the new cycle will be the 06 UTC analysis and forecast in the Boundary Conditions Optional Programme on 11 Jun 2019  followed by the 12 UTC main assimilation and forecast. The monthly forecast extension to the ensemble will be run with the new IFS cycle for the first time the following Thursday 13 Jun 2019.
• We have identified a problem with the parameter "Maximum CAPES in the last 6 hours" (param=228036, shortName=mxcapes6) in Cycle 46r1, affecting the data archived in MARS and available through the dissemination, for HRES, ENS and ENS extended. We are working on a fix for this parameter.

15 May 2019 

• We have now reached the release candidate phase of the implementation of the new IFS Cycle 46r1.
• IFS cycle 46r1 test data is available in dissemination.
• Changes in some Cycle 46r1 data are highlighted.

05 Apr 2019 

• The expected date for the operational implementation of IFS cycle 46r1 is 11 Jun 2019. We will confirm this date early in May.
• The second set of live-streamed seminars for cycle 46r1 will take place on 15 May 2019 at 09:30 BST and on 16 May 2019 at 17:00 BST.

28 Mar 2019

• The IFS Cycle 46r1 test data is available in MARS including new model output parameters.

Cycle 46r1 live-streamed seminars

We have organised two live-streamed seminars to introduce the new IFS cycle 46r1. The first seminar has mainly covered the Scientific changes made in the new Cycle (see recording and slides below). The second seminar focused more on the meteorological impact and scores of the new cycle, with details on how to access the Cycle 46r1 release candidate  test data.

The recording of the first cycle 46r1 seminar is available at https://ecmwf.adobeconnect.com/pgy081jw03ya/. The  presentation slides are also available separately at https://www.ecmwf.int/sites/default/files/medialibrary/2019-02/46r1_overview_AndyBrown.pdf.

The recording and slides of the second  cycle 46r1 seminar are available under https://www.ecmwf.int/en/about/media-centre/news/2019/ecmwf-holds-webinars-june-upgrade-its-forecasting-system.

Timetable for implementation

The planned timetable for the implementation of the cycle 46r1 is as follows:

The timetable represents current expectations and may change in light of actual progress made.

Meteorological content of the new cycle

 Assimilation

• Continuous data assimilation. Number of 4D-Var outer loops increased from 3 to 4. Early delivery assimilation window length increased from 6 to 8 hours. Observation cut off time extended.
• Ensemble of Data Assimilations (EDA) increased from 25 members to 50 members.
• Use of the EDA spread to compute the Simplified Extended Kalman Filter (SEKF) soil analysis Jacobians.
• Weakly coupled data assimilation introduced for sea-surface temperature in the tropics only.
• Consistent spatial interpolation of the model to observation locations in trajectories and minimisations. Interpolation in nonlinear trajectories changed from bicubic to bilinear interpolation.
• RTTOV upgraded from v12.1 to v12.2.

Observations

• Assimilation of SMOS neural network soil moisture product.
• Assimilation of SSMIS-F17 150h GHz and GMI 166 v/h GHz.
• Improved use of land sea mask in the field of view for microwave imagers.
• Introduction of interchannel observation error correlations for ATMS.
• Introduction of interchannel observation error correlations for geostationary water vapour channels.
• Slant path calculations for geostationary radiances.
• Extend usage of geostationary radiances to higher zenith angles.
• Consistent infrared aerosol detection

For further details, read Main Contributions in data assimilation and observations.

Model

• Improvements in convection scheme (entrainment, CAPE closure, shallow convection).
• Activate LW scattering in radiation scheme.
• 3D aerosol climatology replaces 2D climatology.
• Correct scaling of dry mass flux in diffusion scheme.
• Improvement of the TL/AD of the semi-Lagrangian departure point scheme in the polar cap area.
• Fix instability in 2m temperature diagnostic related to wet tile.
• Bug fix in the computation of rain amount that could freeze when intercepted by the snow-pack.
• New parametrisations for wind input and deep water dissipation for for the wave model.
• Limit on wave spectrum for very shallow water and minimum depth set to 3m.
• ENS makes use of 50 EDA-members and initial perturbations are made exchangeable.
• ENS radiation time-step is reduced from 3 hours to 1 hour, to be consistent with HRES.
  

For further details, read Main Contributions in modelling.

Meteorological impact of the new cycle

The following evaluation of the new cycle is based on the alpha and beta testing of the new Cycle. Scorecards for the new IFS Cycle 46r1 are now available.

Weather parameters and waves

The IFS Cycle 46r1 brings substantial improvements in forecast skill both for the ENS and the HRES. Medium-range forecast errors in the extra-tropics are reduced by 1-5% for upper-air and by 0.5-2% for surface parameters. Improvements of this magnitude are seen both against analysis and against observations. In terms of lead time, upper-air improvements amount to a gain of the order of 2-3 hours. In the tropics, HRES results are predominantly positive, but there are some increases in temperature and humidity errors, mainly seen in verification against analysis. For temperature, they are due to changes in the analysis and the introduction of a 3D aerosol climatology. ENS results in the tropics are also mixed, in addition to the already mentioned changes they are affected by a minor reduction in spread on the order of 1% due to changes in the deep convection scheme. Wave parameters (significant wave height and mean wave period) in the HRES are improved by 5-10% due to a major upgrade in the ocean wave model. Increased wave activity leads to some degradation in wave height at longer lead times in the ENS.

Precipitation forecast skill increases in the extra-tropics by about 0.5% in the ENS and 1% in the HRES. Other weather parameters, such as 2m temperature and 2m dewpoint, 10m wind speed, and total cloud cover improve by about 1% in the ENS, and by 0.5-1% in the HRES when verified against observations. In the tropics, slightly reduced spread and increased bias lead to a very small (0.1-0.2%) degradation in ENS precipitation. Scores in the tropics show strong improvements for 2m temperature (4-8% against analyses both in ENS and HRES), (1-2% against obs in ENS).

Tropical cyclones

Results for TCs are generally neutral. There is a slight improvement in the tracks consistent with improvements in tropical winds, but this signal has only marginal statistical significance.

Extended range

The extended-range impact of model changes associated with 46r1 is neutral, except for a small degradation of 2-metre temperature and precipitation skill scores in the tropics. However, the use of ERA5 instead of ERA-Interim as initial condition gives significant improvements in weeks 1-2 in the extratropics, and up to week 4 in the tropics.

Re-forecasts

The new IFS cycle 46r1 will use the ERA5 data to initialize the re-forecasts and also use ERA5 EDA to perturb the re-forecasts initial conditions.

Technical details of the new cycle

Changes to GRIB encoding

Model identifiers

The GRIB model identifiers (generating process identification number) for cycle 46r1 will be changed as follows:

Sea Surface Temperature

New model output parameters

Extended output have been added in cycle 46r1, including a subset of ocean fields on the atmospheric grid.

paramId	shortName	name	Description	units	GRIB edition	Components	Test data available	Dissemination	ecCharts	Added to the Catalogue
Near-surface wind output
228239	200u	200 metre U wind component	eastward component of the 200m wind.	m s-1	1	HRES / ENS		TBC		TBC
228240	200v	200 metre V wind component	northward component of the 200m wind.	m s-1	1	HRES / ENS		TBC		TBC
Wave model parameters
140098	weta	Wave induced mean sea level correction	Wave induced mean sea level correction	m	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
140099	wraf	Ratio of wave angular and frequency width	Ratio of wave angular and frequency width	dimensionless	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
140100	wnslc	Number of events in freak waves statistics	Number of events in freak waves statistics	dimensionless	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
140101	utaua	U-component of atmospheric surface momentum flux	U-component of atmospheric surface momentum flux	N m-2	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
140102	vtaua	V-component of atmospheric surface momentum flux	V-component of atmospheric surface momentum flux	N m-2	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
140103	utauo	U-component of surface momentum flux into ocean	U-component of surface momentum flux into ocean	N m-2	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
140104	vtauo	V-component of surface momentum flux into ocean	V-component of surface momentum flux into ocean	N m-2	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
140105	wphio	Wave turbulent energy flux into ocean	Wave turbulent energy flux into ocean	W m-2	1	HRES-WAM / HRES-SAW / ENS-WAM		TBC		TBC
Ocean parameters available at the surface, produced by the NEMO model.
174098	sithick	Sea-ice thickness *	Sea-ice thickness	m	1	HRES / ENS		TBC		TBC
151148	mld	Mixed layer depth *	Mixed layer depth	m	1	HRES / ENS		TBC		TBC
151145	zos	Sea surface height *	Sea surface height	m	1	HRES / ENS		TBC		TBC
151163	t20d	Depth of 20C isotherm *	Depth of 20C isotherm	m	1	HRES / ENS		TBC		TBC
151130	so	Sea water practical salinity *	Sea water practical salinity	psu	1	HRES / ENS		TBC		TBC
151164	tav300	Average potential temperature in the upper 300m *	Average potential temperature in the upper 300m	degrees C	1	HRES / ENS		TBC		TBC
151175	sav300	Average salinity in the upper 300m *	Average salinity in the upper 300m	psu	1	HRES / ENS		TBC		TBC
* All fields are masked on land and lake points. tav300 and sav300 are masked on ocean points with depth < 300m.
Parameters on Potential Vorticity levels (1.5 and 2 PVU)
129	z	Geopotential	This parameter is the gravitational potential energy of a unit mass, at a particular location, relative to mean sea level.	m2 s-2	1	HRES / ENS *		TBC		TBC
203	o3	Ozone mass mixing ratio	This parameter is the mass of ozone per kilogram of air.	kg kg-1	1	HRES / ENS *		TBC		TBC
3	pt	Potential Temperature	Potential Temperature	K	1	HRES / ENS *		TBC		TBC
54	pres	Pressure	Pressure	Pa	1	HRES / ENS *		TBC		TBC
133	q	Specific humidity	This parameter is the mass of water vapour per kilogram of moist air.	kg kg-1	1	HRES / ENS *		TBC		TBC
131	u	U component of wind	This parameter is the eastward component of the wind. It is the horizontal speed of air moving towards the east, in metres per second. A negative sign thus indicates air movement towards the west.	m s-1	1	HRES / ENS *		TBC		TBC
132	v	V component of wind	This parameter is the northward component of the wind. It is the horizontal speed of air moving towards the north, in metres per second. A negative sign thus indicates air movement towards the south.	m s-1	1	HRES / ENS *		TBC		TBC
* The perturbed forecasts for the ENS only contain the parameters pt, u and v.
Ensemble probabilities
131098	tpg25	Total precipitation of at least 25 mm	Total precipitation of at least 25 mm	%	2	ENS		TBC		TBC
131099	tpg50	Total precipitation of at least 50 mm	Total precipitation of at least 50 mm	%	2	ENS		TBC		TBC
131085	tpg100	Total precipitation of at least 100 mm	Total precipitation of at least 100 mm	%	2	ENS		TBC		TBC
131100	10fgg10	10 metre wind gust of at least 10 m/s	10 metre wind gust of at least 10 m/s	%	2	ENS		TBC		TBC
133093	ptsa_gt_1stdev	Probability of 850hPa temperature standardized anomaly greater than 1 standard deviation	Probability of temperature anomaly greater than 1 standard deviation of the climatology.	%	2	ENS		TBC		TBC
133094	ptsa_gt_1p5stdev	Probability of 850hPa temperature standardized anomaly greater than 1.5 standard deviation	Probability of temperature anomaly greater than 1.5 standard deviation of the climatology.	%	2	ENS		TBC		TBC
133095	ptsa_gt_2stdev	Probability of 850hPa temperature standardized anomaly greater than 2 standard deviation	Probability of temperature anomaly greater than 2 standard deviation of the climatology.	%	2	ENS		TBC		TBC
133096	ptsa_lt_1stdev	Probability of 850hPa temperature standardized anomaly less than -1 standard deviation	Probability of temperature anomaly less than -1 standard deviation of the climatology.	%	2	ENS		TBC		TBC
133097	ptsa_lt_1p5stdev	Probability of 850hPa temperature standardized anomaly less than -1.5 standard deviation	Probability of temperature anomaly less than -1.5 standard deviation of the climatology.	%	2	ENS		TBC		TBC
133098	ptsa_lt_2stdev	Probability of 850hPa temperature standardized anomaly less than -2 standard deviation	Probability of temperature anomaly less than -2 standard deviation of the climatology.	%	2	ENS		TBC		TBC
Ensemble mean / Ensemble standard deviation
10	ws *	The speed of horizontal air movement in metres per second.	The speed of horizontal air movement in metres per second.	m s-1	1	ENS		TBC		TBC
130	t *	This parameter is the temperature in the atmosphere.	This parameter is the temperature in the atmosphere.	K	1	ENS		TBC		TBC
* These parameters have been added at 250 hPa.
Extreme Forecast Index (EFI) & Shift Of Tails (SOT)
132045	wvfi	Water vapour flux index	EFI and SOT for water vapour flux	(-1 to 1)	1	ENS		TBC		TBC
132167	2ti	2 metre temperature index	EFI and SOT for weekly mean temperature (out to week 6)	(-1 to 1)	1	ENS-EXTENDED		TBC		TBC
132228	tpi	Total precipitation index	EFI and SOT for 1 week total precipitation (out to week 6)	(-1 to 1)	1	ENS-EXTENDED		TBC		TBC

Changes in Cycle 46r1 data

EFI/SOT in the extended-range 

With Cycle 46r1, the EFI and SOT become available in the extended-range forecast for two parameters: 7-day mean of 2m temperature and 7-day total precipitation. In contrast to the medium range, the model climate is derived from a set of 3 (rather than 9) re-forecast run dates, centred on the date of the real-time forecast initialisation (all are from 00UTC). The climate sample size is therefore much smaller and comprises 660 values compared to 1980 used in the medium range, but this does make the EFI and SOT consistent with other climate-related products from the extended-range forecasts, such as anomalies and probabilities. 

New climatology for 850hPa Temperature anomaly probabilities 

The outdated fixed climatology for computing historical 850hPa temperature anomaly probability parameters 131020 (talm2), 131021 (tag2), 131022 (talm8), 131023 (talm4), 131024 (tag4) and 131025 (tag8) is replaced by a new re-forecast-based climatology. The same climatology is also used to compute the Cycle 46r1 new (standard-deviation-related) 850hPa temperature anomaly probability parameters listed above. The new climatology is much more compatible with the real-time forecast. 

A change to the computation of maximum CAPE and maximum CAPE-shear parameters 

The two parameters 228035 (mxcape6) and 228036 (mxcapes6) namely “maximum CAPE in the last 6 hours” and “maximum CAPE-shear in the last 6 hours” respectively implemented with cycle 45r1, were computed in a complex way by combining hourly output of the model’s instantaneous CAPE (paramID=59) and CAPE-shear (paramID=228044) with a different type of CAPE, based on virtual temperature, that is used more directly by the model parametrisation where convection is active. This way of computing mxcape6 and mxcapes6 is inconsistent with the standard, instantaneous CAPE and CAPE-shear output fields provided by ECMWF hitherto. So to achieve more consistency, from cycle 46r1 we will change the mxcape6 and mxcapes6 variables to be based solely on the standard instantaneous hourly values of CAPE and CAPE-shear.

Software

ecCodes

ecCodes version 2.12.5 provides full support for the new model output parameters introduced in IFS Cycle 46r1.

Magics

Magics version 4.0.3 provides full support for the new model output parameters introduced in IFS Cycle 46r1.

Metview

Metview version 5.5.3 provides full support for the new model output parameters introduced in IFS Cycle 46r1.

Availability of test data from the IFS cycle 46R1 test suites

Test data in MARS

Test data from the IFS Cycle 46r1 test suites are available in MARS. The data are available with experiment version 0073 (MARS keyword EXPVER=0073) starting from 00 UTC on 29 January 2019.

The data can be accessed in MARS from:

• HRES (class=od, stream=oper, expver=73)
• Wave HRES (class=od, stream=wave, expver=73)
• ENS (class=od, stream=enfo, expver=73)
• ENS Wave (class=od, stream=waef, expver=73)

Only registered users of ECMWF computing systems will be able to access the test data sets in MARS.

The data should not be used for operational forecasting.  Please report any problems you find with this data to Service Desk.

Test data in dissemination

IFS Cycle 46r1 test data from the release candidate testing stage are available through the test dissemination system, starting from the 12Z run on 14 May 2019. Users of ECMWF dissemination products can trigger transmission of test products by logging in to the test ECPDS system at https://ecpds-xmonitor.ecmwf.int/  (or https://msaccess.ecmwf.int:7443) in the usual manner. In order to receive the test products, users have to have their firewall open to the relevant ECPDS Data Movers:

• Internet transfers: 193.61.196.104 ( ecpds-xma.ecmwf.int ), 193.61.196.105 ( ecpds-xmb.ecmwf.int ) and 193.61.196.113 ( ecpds-xmc.ecmwf.int )
  

• RMDCN transfers: 136.156.8.132 ( mspds-dm4.ecmwf.int ) and 136.156.8.133 ( mspds-dm5.ecmwf.int )

The IFS Cycle 46r1 test products are available as version number 73 (file names ending with '73'). The test products are generated shortly behind real-time and based on the operational dissemination requirements and the IFS Cycle 46r1 test data for HRES, HRES-WAM, HRES-SAW., ENS, ENS-WAM and ENS extended.

Should you require any assistance with IFS Cycle 46r1 test dissemination products, please contact Data Services.

Graphical display of IFS cycle 46r1 test data using ecCharts

From the run of 16 May at 00Z onwards, the IFS cycle 46r1 layers are available in ecCharts. Cycle 46r1 layers are identified by the label "0073" in their title.

Web charts based on IFS cycle 46r1 test data

ENS meteograms based on IFS cycle 46r1 test data are available and can be viewed by selecting the "IFS cycle 46r1' model run in the ENS meteograms interface. Access to remaining web charts in the Charts Catalogue will be available shortly.

Time-critical applications

Option 1 - simple time-critical jobs

Member State users of the  "Simple time-critical jobs" framework can test that their scripts will work with the IFS Cycle 46r1 test data by using the limited ECaccess 'events' set up for this purpose:

For these events, MSJ_EXPVER environment variable is set to 0073 and can be used to specify the IFS Cycle 46r1 test data in any MARS retrievals.

These events are  intended for testing technical aspects only and should not be used for Time Critical activities.

Options 2 and 3

Option 2 or 3 time-critical applications can be tested with the IFS Cycle 45r1 test data retrieved from MARS or received in Dissemination.

Further reading

• Peter Lean, Massimo Bonavita, Elías Hólm, Niels Bormann,Tony McNally: "Continuous data assimilation for the IFS" - ECMWF Newsletter 158. Available at https://www.ecmwf.int/en/newsletter/158/meteorology/continuous-data-assimilation-ifs

• Simon Lang, Elías Hólm, Massimo Bonavita, Yannick Trémolet (JCSDA, US): "A 50-member Ensemble of Data Assimilations" - ECMWF Newsletter 158. Available at https://www.ecmwf.int/en/newsletter/158/meteorology/50-member-ensemble-data-assimilations

• ECMWF 's website news item: "Upgrade to boost quality of ocean wave forecasts". Available at https://www.ecmwf.int/en/about/media-centre/news/2019/upgrade-boost-quality-ocean-wave-forecasts
• Frédéric Vitart, Gianpaolo Balsamo, Jean-Raymond Bidlot, Simon Lang, Ivan Tsonevsky, David Richardson, Magdalena Alonso-Balmaseda: "Use of ERA5 to Initialize Ensemble Re-forecasts" - ECMWF Technical Memorandum nr. 841. Available at https://www.ecmwf.int/en/elibrary/18872-use-era5-initialize-ensemble-re-forecasts

Document versions