Page History
...
Term | Meaning | |
---|---|---|
Term | Meaning | |
ECMWF | European Centre for Medium Range Weather Forecasts. | |
RSMC | Regional Specialized Meteorological Centre are set up under the auspices of WMO. RSMCs have ultimate responsibility for official forecasts of tropical cyclones within their respective regions. These are based on interpretation of forecast data from a number of NWP centres (including ECMWF). RSMC bulletins take precedence over predictions from any individual NWP models. | |
Forecaster User Guide FUG | Online user guide aimed primarily for use by forecasters to amplify and explain products from the ECMWF integrated forecasting system. | |
Terms associated with numerical weather prediction models | ||
NWP | numerical weather prediction. The term is used where statements are applicable to numerical forecast models in general, and not only to those in the ECMWF Integrated Forecasting System. | |
IFS | The ECMWF integrated forecasting system is the comprehensive earth-system numerical data assimilation and weather prediction model developed at ECMWF that provides forecasts and associated verification at different resolutions and for several time ranges. Links: FUG | |
LAM | Local area model. A regional or smaller area model with high resolution and using boundary conditions supplied by a larger area model. Several National and other weather forecasting centres use their own local area models with boundary conditions driven by ECMWF IFS model output. | |
HRES medium range forecast | The ECMWF forecast system with high spatial resolution using the ECMWF integrated forecast system for producing deterministic medium range forecasts. It is a land- and sea-coupled, 3-dimensional general circulation atmospheric model run four times per day currently at horizontal resolution 9km and vertical resolution 137 levels. It provides deterministic information on the atmospheric evolution for the 10-day period (day 0 to day 10). With effect from the introduction of Cy48r1 in June 2023 the medium range ensemble has the same horizontal and vertical resolution as HRES and the unperturbed ensemble member (the control) and HRES are almost identical. HRES has been retained in Cy48r1 for continuity purposes. | |
ENS medium range forecast | The ECMWF forecast system with high spatial resolution using the ECMWF integrated forecast system for producing probabilistic medium range forecasts. It is a land- and sea-coupled, 51-member ensemble, 3-dimensional general circulation atmospheric model run four times per day at currently at horizontal resolution 9km and vertical resolution 137 levels. It provides probability information on the atmospheric evolution for the 15-day period (day 0 to day 15). With effect from the introduction of Cy48r1 in June 2023 the medium range ensemble has the same horizontal and vertical resolution as HRES and the unperturbed ensemble member (the control) and HRES are almost identical. Links: FUG | |
extended range | The ECMWF forecast system with moderate spatial resolution using the ECMWF integrated forecast system for producing probabilistic extended range forecasts. It is a land- and sea-coupled, 101-member ensemble, 3-dimensional general circulation atmospheric model run once per day currently at horizontal resolution 36km and vertical resolution 137 levels. It provides probability information on the atmospheric evolution for the the 32-day period (day 15 to day 46). It provides an overview of the atmospheric evolution and focuses mainly on the week-to-week changes. It bridges the gap between medium-range and seasonal forecasting. Links: FUG. | |
seasonal | The ECMWF forecast system with a moderate spatial resolution and a reduced vertical resolution using the ECMWF integrated forecast system for producing probabilistic long-range forecasts. This is a land- and sea-coupled, 51-member ensemble, 3-dimensional general circulation atmospheric model run once per month currently at horizontal resolution 36km and vertical resolution 91 levels). It provides a broad overview of the atmospheric evolution for the period month 2 to month 7, and extended to month 13 four times per year. Links: FUG | |
long-range SEAS5 | Alternative names for the ECMWF seasonal forecast system. | |
ECWAM | The ECMWF ocean wave forecast system with global coverage. This model has 2-way coupling with the atmospheric and ocean models and is run twice per day with resolutions broadly corresponding to the associated atmospheric models. It provides information on the 2-dimensional surface wave spectrum of both oceanic and coastal waters. Links: FUG. | |
HRES-WAM | This deterministic system has 2-way coupling with the HRES atmospheric and ocean models and is run twice per day with spatial resolution broadly corresponding to the HRES forecast system. It provides probability information on the ocean wave evolution for the 10-day period (day 0 to day 10). Links: FUG. Presentation. | |
ENS-WAM | This probabilistic system has 2-way coupling with the ensemble atmospheric and ocean models and is run twice per day with spatial resolution broadly corresponding to the medium range forecast system. It provides probability information on the ocean wave evolution for the 15-day period (day 0 to day 15). Links: FUG. Presentation. Products. | |
extended WAM | This probabilistic system has 2-way coupling with the atmospheric and ocean models and is run twice per week (on Mondays and Thursdays) with spatial resolution broadly corresponding to extended range forecast system. It provides probability information on the ocean wave evolution for the 7-day period (day 15 to day 46). Links: FUG. Tropical storm products. | |
SEAS-WAM | This probabilistic system has 2-way coupling with the atmospheric and ocean models and is run twice per day with spatial; resolution broadly corresponding to the seasonal forecast system. FUG. Tropical storm products. | |
NEMO | The nucleus for European modelling of the ocean is an atmosphere-coupled, 51-member ensemble, 3-dimensional general circulation system run with moderate resolution (currently horizontal resolution 0.25° and vertical resolution 75 levels). It provides information on the evolution of sea temperature structure and ocean currents for use with HRES, medium range ENS, extended ENS, and long-range atmospheric models. Links: FUG. | |
LIM2 | The Louvain-la-Neuve sea ice model (Version 2). Itis anumerical model of sea icedesignedfor climate studies and operational oceanography (Centro Euro-Mediterraneo sui Cambiamenti Climatici). It forecasts formation, thickening, transport and melting of sea ice. it is coupled tothe oceangeneral circulation model and is part of NEMO. Links: CMCC | |
LIM3 | The Louvain-la-Neuve sea ice model (Version 3). It is more advanced but is not currently used by ECMWF. Links: CMCC. Lim3 Documentation. | |
EUROSIP | An discontinued multi-model ensemble of international seasonal forecasts giving ensemble mean anomalies of a limited number of monthly mean fields. It was discontinued in 2019 and is superseded by the Copernicus Climate Change Service (C3S). Links: FUG. ECMWF description | |
C3S | The Copernicus Climate Change Service (C3S) combines observations of the climate system with the latest science to develop authoritative, quality-assured information about the past, current and future states of the climate in Europe and worldwide. Links: Copernicus climate change services | |
Terms associated with ensemble forecasting | ||
analysis | A detailed estimate of the current state of the atmosphere computed using, as optimally as possible, observations and prior information about the Earth-system using ECMWF’s highest resolution model. | |
ensemble | A set of numerical forecasts. In order to estimate the effect of possible initial analysis errors and the consequent uncertainty of the forecasts, an ensemble is formed of many (currently 50 in medium range ensemble, 100 in extended range ensemble) different initial states (the perturbed members) and one unperturbed initial state (the control member). The distribution and spread of these results gives an indication of the predictability and confidence of the evolution of the forecast. Links: FUG | |
perturbation | Adjusted values introduced into each ensemble member to give a set of alternative analyses and forecasts with each member being truly independent of all the others. They derived from the ensemble of data assimilations and singular vectors. Link: FUG | |
control member (CNTL) | The control member is the unperturbed member of an ensemble of numerical forecasts. It uses the most accurate estimate of the current conditions and the currently best description of the model physics. The perturbations for the remainder of the ensemble members are applied to the control analysis. | |
ensemble mean (EM) | The ensemble mean of an ensemble of forecasts is the average value over all the results from ensemble members for a given time. | |
ensemble median | The ensemble median of climate data or an ensemble of forecasts is the value where half of the ensemble member values lie below and half of the ensemble member values lie above this value. It is the middle value of all the values given by ensemble members. It is not the same as the ensemble mean nor the mid-point of the range of values. Link: FUG | |
probability | Probability in ensemble forecasting is considered to be the proportion of ensemble members showing given characteristics in their individual forecasts. Thus, for a given place and time, if 17 members of a 51 member ensemble have forecast a temperature below 0°C then it is taken that there is a 33% probability of this actually happening. Links: FUG | |
tercile, quintile, decile | Partitioningof ensemble forecast or climate data, ordered from lowest to highest, into groups such that each contains one-third (tercile), one-fifth (quintile), one-tenth (decile) of the total data. Thus, for example, the upper quintile will contain the highest 20% of the data. | |
ensemble spread | This is taken as the standard deviation of results from an ensemble of forecasts. Lower spread implies higher confidence in results. Link: FUG | |
weight | The weight that may be put on each model output to reflect its likely accuracy. | |
clustering | A method of grouping height patterns as forecasted by ensemble members according to some measure. This simplifies user-analysis of ensemble output by identifying possible forecast scenarios and particularly by highlighting the more predictable or probable model forecast patterns. Links: Medium range clustering. Extended range clustering. | |
regime | A method of grouping height patterns as forecasted by ensemble members according to their similarity to defined regimes (North Atlantic Oscillation , NAO-, Block, Atlantic Ridge (ATR), and No Regime). Link: FUG. | |
CDF | The cumulative distribution function is the probability that a result of an individual ensemble member forecast has a value less than or equal to particular values (e.g. a given temperature). Links: FUG. | |
The probability density function shows the probability that results of ensemble member forecasts have a value equal to given values (e.g. temperature). It shows whether results of an ensemble forecast are grouped around a particular value (e.g. a given temperature). Links: FUG | ||
extreme forecast index, EFI | ECMWF has developed the extreme forecast index to help indicate the potential for extreme weather conditions by showing how extreme the ensemble forecast is compared with the model climate (M-climate). It is computed from the difference between cumulative distribution function curves of the the current ENS forecast distribution and the M-climate. The calculations are made so that more weight is given to differences in the tails of the (climatological) distribution in order to better identify extremes. Links: FUG | |
shift of tails, SOT | ECMWF has developed the shift of tails index to help indicate the potential for extreme weather conditions by comparing the tails (extremities) of the ensemble forecasts and the tails of the M-climate distributions and gives information about how extreme an event could potentially be. Links: FUG | |
probability of precipitation, PoP | Probability of precipitation is a measure of the probability that at least a minimum defined quantity of precipitation (which includes: rain, sleet, wet snow, dry snow, freezing rain and ice pellets but not graupel or hail) will occur within a specified forecast period at given location. | |
model output statistics, MOS | Model output statistics is a post-process technique which calibrates deterministic forecast results against a long record of verifying observations using a linear regression scheme. Links: FUG | |
ensemble model output statistics, EMOS | Ensemble model output statistics is similar to model output statistics adapted for use with ensembles. Links: FUG | |
Terms associated with surface energy exchanges | ||
HTESSEL | The ECMWF scheme using a system of tiles which represent the land characteristics within each grid box and describes the evolution of soil, vegetation and snow conditions over the land at various spatial resolutions. It is used operationally to describe energy, temperature and moisture exchanges between the model atmosphere and land and also incorporates land surface hydrology. Links: FUG | |
FLake | The ECMWF scheme using a system of tiles which represent the water characteristics within each grid box and describes the evolution of surface temperature conditions over water at various spatial resolutions. It is used operationally to describe energy, temperature and moisture exchanges between the model atmosphere and lakes and coastal waters. Links: FUG | |
leaf area index, LAI | The leaf area index represents the proportion of leaf area at the earth's surface within each grid box and determines the degree of evapotranspiration exchange with the model atmosphere. Leaf area index varies climatologically month by month order to represent to the typical changes in local vegetation but does not represent any anomalous differences that might occur. Links: FUG | |
multi-layer snow model | The IFS multi-layer snow model uses up to five layers to represent the snowpack and the complex heat fluxes and interactions between them. It represents the vertical structure and evolution of snow temperature, snow mass, density, and liquid water content in each layer. Links: FUG. | |
albedo | Albedo is the fraction of solar energy (shortwave radiation) reflected from the earth back into space. Links: FUG. | |
land-sea mask, LSM | The land-sea mask represents the proportion of land or sea within each grid box and is used to determine the proportional energy, temperature and moisture exchanges between the model atmosphere and the underlying surfaces (using HTESSEL and FLake). Links: FUG | |
grid box | The octahedral Gaussian grid is used for execution of the models. A grid box is the triangular area between the grid points. Link: FUG The grid box used for Meteograms is four grid points surrounding the location in question. Link: FUG | |
CONV | ECMWF integrated forecasting system parametrization of convection. | |
LSWP | ECMWF integrated forecasting system parametrization of large scale water processes. | |
TGWD | ECMWF integrated forecasting system parametrization of turbulence & gravity wave drag. | |
NOGW | ECMWF integrated forecasting system parametrization of non-orographic gravity wave drag. | |
Terms associated with model climates | ||
M-climate | The model climate. A "climate" relevant to the HRES and ENS range based on re-forecasts of 20 years data using the same up-to-date model structure and resolution. All parts of the globe are covered. A model climate is not the same a climate from observed data, although for most aspects they are closely correlated. Links: FUG | |
ER-M-climate | The extended range model climate. A "climate" relevant to the extended range based on re-forecasts of 20 years previous data using the same up-to-date model structure and resolution. All parts of the globe are covered. A model climate is not the same a climate from observed data, although for most aspects they are closely correlated. Links: FUG | |
S-M-climate | The seasonal model climate. A "climate" relevant to the seasonal range based on a reanalyis of 30 years previous data using the same up-to-date model structure and resolution. All parts of the globe are covered. A model climate is not the same a climate from observed data, although for most aspects they are closely correlated. Links: FUG | |
Terms associated with data assimilation | ||
4D-Var | A data assimilation system that iteratively adjusts the initial conditions of a short-range forecast to bring it into closer agreement with meteorological observations in space and time. The adjustments are made in a manner which respects the physical laws that govern the Earth system. The adjusted forecast then serves to define the initial conditions of a new forecast. Links: FUG | |
3D-Var | A data assimilation system that iteratively adjusts the initial conditions of a short-range forecast to bring it into closer agreement with meteorological observations in the assimilation interval and weighted according to their distance in time from the nominal data time, while at the same time tries to stay as close as possible to the first guess. It is a simpler and less effective technique than 4D-Var but is less expensive in computer time. | |
NEMOVAR | Multi incremental and multivariate variational data assimilation system for the NEMO ocean model. It produces ensembles of data assimilation for the ocean model ensemble. Links: FUG | |
Terms associated with initialisation of ensembles | ||
ensemble of data Assimilations, EDA | The ensemble of data assimilation provides an ensemble of estimates of the current state of the atmosphere and its uncertainty. The EDA estimate of the analysis uncertainty can be used as an approximation of the 4DVAR uncertainty. Links: FUG | |
singular vectors, SV | Singular vectors are vector additions to information (the perturbations) within the IFS system which optimally instigate maximum growth rate of those perturbations. Links: FUG | |
EDA SVINI | This is a combination of EDA- and SV-based perturbations. The aim is to provide a better tuned and more skilful ensemble system for the entire forecast range and for the whole vertical structure of the atmosphere. Links: ECMWF description | |
SKEB | A stochastic backscatter scheme. Representation of uncertainty in model energy transfer from sub-grid scales to resolved scales. Not used after Cycle 43R3. | |
SPPT | A stochastically perturbed parametrization tendencies scheme yielding a presentation of uncertainty in model parametrisation schemes and initial conditions. Links: FUG | |
Terms associated with data assimilation and creation of model climates - Atmospheric models | ||
reanalysis | Reanalysis data are a blend of observations with past short-range weather forecasts which are rerun with modern weather forecasting models. Reanalysis data provide the most complete picture currently possible of past weather and climate. They are globally complete and consistent in time. Reanalysis data are used for monitoring climate change, for research and education, for the initialisation of re-forecasts, and for commercial applications. Links: FUG | |
re-forecasts | Re-forecasts are forecasts run using the same model version as the real-time forecast for a number of past dates and can be used to calibrate real-time ensemble forecasts. Re-forecasts of past data to provide a uniform global coverage of a "climate" as defined by the model. It captures the extremes of observed past conditions relevant to the period of the forecast, within the constraints of the model resolution. A model climate is not the same a climate from observed data, although for most aspects they are closely correlated. Re-forecasts are used extensively to enable construction of a model climate (e.g. M-climate). Links: FUG | |
ERA5 | A global reanalysis based on data from 1950 to the present and updated in near real time. It provides hourly estimates of a large number of atmospheric, land and oceanic climate variables with 31km resolution using 137 levels. Reanalyses (and re-forecasts) of past data are used, in part, to provide a uniform global coverage of "climate" as defined by the model. It also captures the extremes of observed past conditions, within the constraints of the model resolution. ERA5 provides information about uncertainties for all variables at reduced spatial and temporal resolutions and is used extensively to enable construction of a model climate. Links: FUG, ECMWF documentation | |
ERA-40 | A global climate reanalysis based on data from 1957 to 2002. Now not used. Links: ECMWF documentation | |
ERA-Interim | A global climate reanalysis from 1979 to 2019 superseded by ERA5. Re-analyses (and re-forecasts) of past data that are used, in part, to provide a uniform global coverage of "climate" as defined by the model. It also captures the extremities of observed past conditions, within the constraints of the model resolution (80km). Now not used. Links: ECMWF description | |
CERA-20C | A global 10-member ensemble of coupled climate reanalyses between 1901 and 2010. It is based on the CERA assimilation system, which assimilates only surface pressure, marine wind observations, and ocean temperature and salinity profiles. Now not used. Links: ECMWF description | |
CAMS | Copernicus atmosphere monitoring service. This monitors aerosols, ozone and other reactive gases, dust, sand, smoke and volcanic aerosols. Links: Copernicus Atmosphere Monitoring Service | |
LDAS | The land surface model analysis includes the analyses of snow depth, soil moisture, soil temperature, snow temperature, and screen-level parameters. Links: ECMWF description | |
RTTOV | A radiative transfer model for the assimilation of satellite radiance observations from the TIROS operational vertical sounder (TOVS). The model deals with evaluation of incoming and outgoing radiation through the model atmosphere and involves interactions with atmospheric constituents (gas molecules, aerosols, clouds, hydrometeors) and also the surface. Link: ECMWF Tech Memo | |
Terms associated with data assimilation and creation of model climates - Ocean models | ||
OCEAN5 | OCEAN5 data is made up of the historical ocean reanalysis (ORAS5) and the daily real time ocean analysis (ORTA5). It uses the same ocean and sea-ice model components as the coupled forecasts in the seasonal atmospheric model. OCEAN5 is essentially associated with seasonal atmospheric model because changes are slow. Link: ECMWF Tech Memo, Ocean analysis/reanalysis | |
ERA-interim wave | A global wave model climate reanalysis from 1979 to the present, updated in near real time. | |
ORAS5 | A global ocean reanalysis based on data from 1979 to the present and is updated with only a few days delay. It provides an estimate of the historical ocean state. Link: ECMWF Tech Memo 823. Ocean analysis/reanalysis | |
ORAS4 | Global ocean reanalysis based on data from 1958 to 2015 superseded by ORAS5. | |
ORTA5 | A global ocean real-time analysis system (version 5) provides an estimate of the ocean and sea-ice initial state and its uncertainty for all ECMWF coupled forecasting configurations. | |
ORTA4 | Global ocean real-time analysis system (version 4) superseded by ORTA5. | |
OSTIA | Operational sea surface temperature and sea ice analysis. Link: Met Office OSTIA | |
Terms associated with convection modelling | ||
CAPE | Convective available potential energy. The energy available to a parcel of air rising under free convection. Indicative of existence of conditional instability and availability of moisture at the base of the ferry convection. CAPE is an indicator of the potential for hazardous weather. Link: FUG, product description | |
CAPE-shear, | The product of CAPE and the vertical wind shear between 925hPa and 500hPa (i.e. the lower troposphere). CAPE-shear is an indicator of the potential for severe storms and hazardous weather. Link: FUG, product description | |
CAPESHEAR, CAPES, CSP | Alternative terms for CAPE-shear used occasionally elsewhere in non-ECMWF literature. | |
CIN | Convective inhibition. The amount of energy that is needed to raise a parcel of air from a given level (such as the surface) up to the level of free convection arising from parcel ascent from that selected level. Link: FUG, product description | |
level of free convection, LFC | The height at which a parcel of air, when lifted, becomes warmer than its surroundings and thus convectively buoyant. | |
lifted condensation level, LCL | The height at which the relative humidity of a parcel of air parcel will reach 100% with respect to liquid water when it is cooled by dry adiabatic lifting. | |
instability index | Instability indices are a way of indicating the potential for convection using simple formulae based upon temperature and moisture data observed or forecast at a small number of pressure levels. Several of these formulae have been derived semi-empirically. Instability diagnostics are better served by CAPE-related variable. Links: FUG | |
Terms associated with non-convection modelling | ||
large scale precipitation | Large scale precipitation refers to precipitation developed by non-convective processes. | |
total precipitation | Total precipitation is the total of large scale and convective precipitation. | |
Terms associated with parametrization | ||
CONV | ECMWF integrated forecasting system parametrization of convection. | |
LSWP | ECMWF integrated forecasting system parametrization of large scale water processes. | |
NOGW | ECMWF integrated forecasting system parametrization of non-orographic gravity wave drag. | |
TGWD | ECMWF integrated forecasting system parametrization of turbulence and gravity wave drag. | |
sub-grid scale drag | A roughness length parametrization is used to obtain a realistic area-averaged turbulent drag according to the underlying surface (e.g. deserts, grass, forests). An orographic parameter is derived from the height of valleys, hills and mountains to represent orographic features that are too small to be resolved by the model grid. | |
Terms associated with ocean forecasting | ||
wave direction | Wave direction is described as the direction the waves are moving towards. This is opposite from the convention for wind direction which is defined by where the winds are coming from. Links: FUG | |
wave height | Wave height is defined as the vertical distance between trough and crest. Links: FUG | |
significant wave height SWH, H1/3 or Hs | The significant wave height represents the average height of the highest third of surface ocean waves. Links: FUG | |
wave spectrum | The irregular surface of the sea is a combination of waves with different heights, lengths and directions. The wave spectrum describes distribution of energy among these wave components. Links: FUG | |
wind-sea, wind waves | Wind-sea is the part of the wave spectrum directly affected by local winds. Links: FUG | |
swell | Swell is the part of the wave spectrum that is not associated with the local wind but were generated by the wind at a different location and time. Links: FUG | |
shoaling | Shoaling is the deformation of waves moving from the ocean into shallow waters causing the waves to become steeper, increase in height, and have shorter wavelength. Links: FUG | |
Terms associated with meteorological phenomenon | ||
Madden Julian Oscillation, MJO | Madden-Julian Oscillation is a broad-scale wave-like convective phenomena centred on the equator. It is important for tropical predictability on the monthly time scale and interaction with the general atmospheric circulation. It is observed mainly in a sector spanning the Indian and Pacific oceans. Links: FUG, ECMWF paper | |
El Nino | El Niño is used to describe a warmer than average sea surface temperature in the equatorial Pacific and is associated with warmer, wetter than average weather in the tropical eastern Pacific. Links: FUG | |
La Nina | La Niña is used to describe cooler than average sea surface temperature in the equatorial Pacific and is associated with cooler, drier than average weather in the tropical eastern Pacific and wetter weather in the tropical western Pacific. Links: FUG | |
ENSO | El Niño Southern Oscillation refers to the general atmospheric pressure differences between the east and west tropical Pacific that accompany both El Niño and La Niña episodes. The consequent interaction between atmosphere and ocean have a significant influence on later weather in other parts of the world. Links: FUG | |
Terms associated with meteorological regimes | ||
regime | Regimes are recurrent large-scale flow patterns often with a persistence of a week to a month. Each regime allows a broad description of the predominant weather scenario while allowing intuitive interpretation of weather variability. The persistence of weather regimes and their sensitivity to external forcings give rise to increased predictability at the extended range. Links: FUG, ECMWF description | |
NAO regime | North Atlantic oscillation refers to the general atmospheric pressure differences and associated westerly winds between northern and southern areas of the North Atlantic. | |
blocking regime, BLO | Blocking regime refers to the general positioning of atmospheric pressure ridging over the Atlantic or Europe. | |
Atlantic ridge regime (ATR) anti-blocking regime | The terms Atlantic Ridge Regime and Anti-Blocking Regime are used synonymously with atmospheric pressure ridging over the Atlantic | |
NAO-BLO phase space diagram | NAO and BLO circulation systems can be considered as orthogonal so a NAO-BLO phase space diagram may be used to investigate and illustrate the relationship between circulation type and other forecast or observed parameters. NAO-BLO diagrams may be used to illustrate the sequence of transitions from one regime type to another. Ensemble results can be shown as individual plots for a given forecast time or in probably density form regarding results over a few days. Links: FUG | |
ensemble trajectories regime projection | An ensemble trajectory diagram is a NAO-BLO phase space diagram which illustrates the time evolution and strength of the atmospheric regimes forecast by each of the ensemble members. Links: FUG | |
regime charts | Regime charts show, in histogram format, how the probabilities of each of the regimes evolve on a daily basis as the ensemble members progress through the extended range forecast period. Links: FUG | |
Terms associated with verification statistics | ||
ME | Mean error. | |
MAE | Mean absolute error. | |
MSE | Mean square error. | |
RMSE | Root mean square error. | |
ESL | Error saturation level. Model forecast errors do not grow indefinitely but asymptotically approach a maximum about 40% greater than that of a forecast based solely on a climatological average. Links: FUG | |
hit | A hit is a correct forecast that an event will happen and it does actually happen. | |
miss | A miss is an incorrect forecast that an event will not happen but it does actually happen. | |
false alarm | A false alarm is an incorrect forecast that an event will happen but it does not actually happen. | |
correct null | A correct no forecast its a correct forecast that an event will not happen and it does not actually happen. | |
contingency table | A contingency table tabulates the number of hits, false alarms, misses, and correct nulls. Links: FUG | |
HR | Hit rate. The proportion of correct forecasts that an event will happen compared with all the events that actually did occur (i.e. the proportion of hits given the event was observed). | |
FR | False alarm rate. The proportion of incorrect forecasts that an event will happen compared with all the events that actually did not occur (i.e. the proportion of false alarms given the event was not observed.). | |
FAR | False alarm ratio. The proportion of incorrect forecasts that an event will happen compared with all the forecasts that the event will happen (i.e. the proportion of false alarms, given the event was forecast). | |
BS | The Brier score is a measure of how good forecasts are in matching an analysis or a climatology (i.e. M-climate, ER-climate and S-M-climate). Links: FUG | |
BSS | The Brier skill score is a measure of the relative skill of a verified forecast against a reference unskilled forecast. Climatology (i.e. M-climate, ER-climate and S-M-climate) is used as the reference forecast at ECMWF. Links: FUG | |
RPS | The (Discrete) ranked probability score is a measure of forecast values being mis-assigned to a category (e.g. tercile, quintile, etc.) against the corresponding observations. The words "discrete" and "ranked" refer to the discrete nature of the categories. Links: FUG | |
CRPS | The continuous ranked probability score compares the probability distribution of the quantity forecasted by the ensemble forecast system to its observed value. This is a generalisation of ranked probability score (RPS) using categories that are continuous rather than discrete. Links: FUG | |
RPSS | The ranked probability skill score (de-biased) (RPSS-D) compares the RPS of a verified forecast against a reference unskilled forecast. Climatology (i.e. M-climate, ER-climate and S-M-climate) is used as the reference forecast. Links: FUG | |
CRPSS | The continuous ranked probability skill score compares the CRPS of a verified forecast against a reference unskilled forecast. Climatology (i.e. M-climate, ER-climate and S-M-climate) is used as the reference forecast at ECMWF. Links: FUG | |
anomaly | The anomaly is the difference between two values or fields of values for the same location and forecast time. At ECMWF it is usually taken as the difference between the forecast of a value and a reference forecast or climatology (i.e. M-climate, ER-climate and S-M-climate). | |
ACC | The anomaly correlation coefficient measures the spatial correlation between a forecast anomaly from climatology (i.e. M-climate, ER-climate and S-M-climate) and a verifying analysis anomaly from climatology (i.e. M-climate, ER-climate and S-M-climate). ACC represents a measure of how well the forecast anomalies have represented the observed anomalies. Links: FUG | |
ETS | The equitable threat score (same as Gilbert skill score) measures the skill of a forecast relative to chance. It is computed for each station over the given period, and then averaged over an area. It is often used in the verification of precipitation forecasts. Links: FUG | |
SEEPS | Stable equitable error in probability space. Skill measure used as an ECMWF headline score, to monitor long-term trends in performance in forecasting three categories of precipitation (dry, light and heavy) accumulated over 24 hours evaluated against observed precipitation amounts. Links: FUG | |
sharpness | Sharpness is the ability of a forecast system to forecast with high or low probabilities (in effect approaching a yes/no forecast and implying high confidence may be placed in the model) rather than forecasts clustered around a climate value (in effect giving no definite guidance and hence lower confidence in the model). Links: FUG | |
reliability | Reliability is the ability of a forecast system to forecast with realistic probabilities such that forecast probability of an event matches the climatological probability of that event. The reliability can be improved by calibration against verification statistics. Links: FUG | |
discrimination | Discrimination is the comparison within a large sample between the distribution of observations when the forecast system predicts an event to occur with given probability. Good discrimination is where distributions are near coincidence with minimal proportion of false alarm forecasts. Links: FUG | |
ROC ROC diagram | Relative operating characteristics is a measure of the performance of probabilistic forecast systems. Probabilistic forecasts are transformed into yes/no forecasts defined by thresholds varying from 0% to 100% and the hit rates and false alarm rates may be plotted on a ROC diagram. The area beneath the plotted curve gives a measure of the effectiveness of the probabilistic forecast. Links: FUG | |
rank histogram Talagrand diagram | The rank histogram shows whether the verifying observation is statistically indistinguishable from the ensemble member forecasts. It measures consistency (a tendency for a consistent positive or negative bias in the forecast values) and reliability (a bias towards over- or under-confidence). Links: FUG | |
Terms associated with Mean and Spread Charts | ||
Std | Standard deviation used as a measure of the spread of results from an ensemble forecast. Links: FUG | |
MStd | Mean standard deviation is used as a measure of the mean spread of the 30 most recent 00UTC or 12UTC model runs. Links: FUG | |
NStd | Normalised standard deviation is used as a measure of the spread of the latest ensemble normalised by the mean spread of the 30 most recent 00UTC or 12UTC model runs. The normalised spread shows the increase or decrease (but not the magnitude) of the spread over the last 30 days and highlights relatively low or relatively high uncertainty, not the uncertainty itself. Links: FUG | |
Terms associated with cost/benefit aspects and analyses | ||
cost/loss | Cost/loss is the ratio of the cost of taking action (e.g. buying insurance against a forecast event) against the potential loss should an event occur. FUG. | |
cost/benefit | Cost/benefit compares the cost/loss ratio based on a model forecast against Cost/Loss ratio based on an ideal perfect (always correct) forecast. FUG. | |
mean expense | The mean expense is the average loss due to no action being taken because of a forecast. | |
value | Value is the reduction of mean expense of using a forecast system against the mean expense of using climatology. | |
Terms associated with delivery of current ECMWF products | ||
open access charts | A catalogue of ECMWF forecast charts of showing surface, upper air and ocean products which are both free to the user and and accessible to all. These include medium-range, extended-range and long-range forecast charts and also include probability information based on the ECMWF forecast ensemble. | |
ecCharts | A suite of web-based applications to inspect, explore, compare, and visualise ECMWF forecast data in an interactive way. A wide selection of layers representing surface and upper-air parameters from ensemble, high-resolution, extended-range, seasonal and wave forecasts which may be superimposed to aid understanding of the forecast structure of the atmosphere and to assess uncertainty. | |
dashboard | A personal "dashboard space" for logged-in ECMWF users. Regularly-used charts and diagrams to allow easy recall and provide a powerful and flexible forecaster-oriented tool. | |
Terms associated with presentation of probability data | ||
box and whisker | Box and whisker format displays the percentiles of the probability distribution function of a variable as forecast by the ensemble members. The limits of the whiskers correspond to the 5th and 95th percentiles, the limits of the box correspond to the lower and upper tercile, and the median is represented by the line within box. Links: FUG | |
meteogram | A meteogram is a probabilistic representation of the time evolution of the distribution of cloud, precipitation, wind and temperature given by the ensemble model forecasts for a selected location. The representation is in box and whisker format. Links: FUG | |
wavegram | A wavegram is a probabilistic representation of the time evolution of the distribution of wind and wave parameters given by the ensemble model forecasts for a selected location. The representation is in box and whisker format. Links: FUG | |
climagram | A climagram is a probabilistic representation of the time evolution of temperature or precipitation given by the seasonal model forecasts for a selected region. Also shown are the corresponding percentiles of the model climatology (S-M-climate) and observed climatology (from past analyses). The representation is in box and whisker format. Links: FUG | |
plume | A plume is a probabilistic representation in graphical form of the time evolution of the distribution of 850hPa temperature, precipitation, and 500hPa height given by the ensemble model forecasts for a selected location. The graph shows results from high resolution forecast (HRES), the ensemble control (CTRL), together with ensemble members, binned in 12.5% intervals (shaded) and the median. Links: FUG | |
precipitation type histogram | This is a type of precipitation meteogram which gives a probabilistic representation of each type of precipitation given by the ensemble model forecasts for a selected location and forecast time. The representation is in cumulative histogram format. Links: FUG | |
Nino plumes | Nino plumes are a graphical representation of the time evolution of sea surface temperature anomalies given by the seasonal model forecasts for a selected traditional Nino regions of the equatorial Pacific. Link: FUG | |
Terms associated with ECMWF data processing | ||
octahedral reduced Gaussian grid | The octahedral reduced gaussian model grid is a consistent triangular grid spacing across the world, even towards the poles. Links: FUG | |
soil moisture | Soil moisture is a measure of the water content within the ground and is commonly expressed as a percentage of the water that the ground could hold when fully saturated. Links: FUG | |
moist physics | The representation of the cloudy convective atmosphere in the IFS, including moist convective mass flux in the moist convection scheme and dry mass flux and turbulent diffusion in the turbulence scheme. Links: FUG | |
Terms associated with convection modelling | ||
convective precipitation | Convective precipitation is associated with precipitation developed only by convective processes. Links: FUG, Product description | |
CAPE | Convective available potential energy. The energy available to a parcel of air rising under free convection. Indicative of existence of conditional instability and availability of moisture at the base of the ferry convection. CAPE is an indicator of the potential for hazardous weather. Link: FUG, Product description | |
CAPE-shear, | The product of CAPE and the vertical wind shear between 925hPa and 500hPa (i.e. the lower troposphere). CAPE-shear is an indicator of the potential for severe storms and hazardous weather. Link: FUG, Product description | |
CAPESHEAR, CAPES, CSP | Alternative for CAPE-shear used occasionally elsewhere in non-ECMWF literature. | |
CIN | Convective inhibition. The amount of energy that is needed to raise a parcel of air from a given level (such as the surface) up to the level of free convection arising from parcel ascent from that selected level. Link: FUG, Product description | |
level of free convection, LFC | The height at which a parcel of air, when lifted, becomes warmer than its surroundings and thus convectively buoyant. | |
lifted condensation level, LCL | The height at which the relative humidity of a parcel of air parcel will reach 100% with respect to liquid water when it is cooled by dry adiabatic lifting. | |
instability index | Instability indices are a way of indicating the potential for convection using simple formulae based upon temperature and moisture data observed or forecast at a small number of pressure levels. Several of these formulae have been derived semi-empirically. Instability diagnostics are better served by a CAPE-related variable. Links: FUG | |
Terms associated with non-convection modelling | ||
large scale precipitation | Large scale precipitation is associated with precipitation developed by non-convective processes. Links: FUG, Product description | |
LSWP | ECMWF Integrated Forecasting System parametrization of Large Scale Water Processes. | |
point rainfall | Point rainfall is an experimental post-processed product, based on calibrating ENS fields to give a probabilistic forecast of rainfall totals for points within a grid box. Links: FUG | |
total precipitation | Total precipitation is the total of large scale and convective precipitation. Links: Product description | |
Terms associated with ECMWF data archival and retrieval | ||
MIR | ECMWF meteorological interpolation and regridding re-gridding package which can present model or other data graphically on any grid pattern, including 3-D presentations, and include rotation and cropping as desired. Links: FUG, ECMWF MIR/MARS information | |
MARS | ECMWF meteorological archival and retrieval system. Links: ECMWF MARS documentation | |
FDB | ECMWF fields database used for archival and retrieval purposes. | |
Terms associated with data analysis | ||
SCDA short cut-off data assimilation | The early delivery analysis has an early cut-off of incoming data but can nevertheless have sufficient data to provide sufficient information to enable ECMWF forecasts to be delivered without significant loss of effectiveness. Links: FUG | |
early delivery analysis | Alternative name for the short cut-off analysis. | |
LWDA long window data assimilation | The long window data analysis is a stand-alone system for maximum use of available data but analyses are delayed significantly from the nominal data time. Most IFS forecasts use background fields and short cut-off analyses. Links: FUG | |
CoDA | Continuous data assimilation captures delayed data for introduction into the 4D-Var assimilation. This occurs after the calculations have started and allows a more accurate final assimilation and analysis. Links: FUG | |
analysis increments | Analysis increments are significant differences between the observed data and background fields and point out particular areas where the model expectations may be incorrect. Links: FUG | |
General terms | ||
SST | Sea-surface temperature. | |
LSWT | Lake surface water temperature. | |
MSL | Mean sea level. | |
OLR | Outgoing long-wave radiation. | |
SLW | Supercooled liquid water. | |
Terms associated with data interpretation | ||
jumpiness | Jumpiness describes occurrence of forecast results which vary significantly from run to run. Links: FUG | |
forecast jump | Describes a distinct and sudden change in forecast values at a given location and verifying time in a sequence of model runs. Links: FUG | |
forecast flip-flop | Describes alternate higher and lower forecast values at a given location and verifying time in a sequence of model runs. Links: FUG | |
forecast trend | Describes a steady or unsteady move towards a lower or higher forecast values at a given location and verifying time in a sequence of model runs. Links: FUG | |
Terms associated with forecast data presentation | ||
lightning flash density | Lightning flash density is defined as the average or instantaneous number of forecast lightning flashes within a defined area during a defined period. Probability of lightning flash density is defined as the probability that the lightning flash density exceeds a certain density threshold during a defined period. Links: FUG, Product description | |
vertical profile | The vertical profiles forecast product provides information about the vertical structure of the forecast model atmosphere in graphical format (similar to a tephigram, emagram etc) for any location and time. Links: FUG | |
visibility | The ECMWF visibility output is an experimental product calculated using an optical scattering law with forecast surface layer moisture. Correctly forecasting visibility is a major challenge and the product should be used with caution. Links: FUG, Product description | |
types of precipitation | A range of types of precipitation at the surface can be diagnosed from the profiles of rain and snow combined with near-surface model temperature. Types of precipitation are rain, sleet, wet snow, dry snow, freezing rain and ice pellets. Graupel and hail, which are both convective in nature, are not predicted. Links: FUG | |
Terms associated with tropical and extra-tropical cyclones | ||
extra-tropical cyclones | An extra-tropical cyclone is large low-pressure weather system with clouds, rain and sometimes strong winds which occur at mid-latitudes (30° – 60° from the equator). They are not the same as tropical cyclones. | |
extra-tropical cyclone diagrams dalmation plots | These are post-processed ensemble products which represent the location and behaviour of near-surface synoptic-scale features typically associated with adverse weather. Links: FUG | |
tropical cyclone, TC | A tropical cyclone is a low pressure system over tropical or sub-tropical waters, with thunderstorm activity and circulating often very strong low level winds. The term Includes hurricanes and typhoons. | |
tropical cyclone Strike Probability | Strike probability is defined as the proportion of ensemble members that predict that the tropical cyclone will pass within a 120km radius of a given location at any time during the next five days. Links: FUG | |
tropical cyclone Lagrangian meteograms | Lagrangian meteograms show, in box and whisker format, a time series of the central pressure and of the 10m wind speed maximum forecasted within a 7ºx7º lat-long box centred on the cyclone and following its motion in each forecast member. Links: FUG | |
tropical storm frequency | This is the frequency of tropical storms predicted by extended range model (for weekly periods) or seasonal model (for six-month periods) averaged over an ocean basin. The results are normalised against the frequency of tropical cyclones identified by the re-forecasts to allow easy comparison with model climate data. Links: FUG | |
accumulated cyclone energy (ACE) | The accumulated cyclone energy is the totalled energy, taking into account number, strength, and duration, of all the tropical storms predicted by extended range model (for weekly periods) or seasonal model (for six-month periods) over an ocean basin. The results are compared with a similarly derived accumulated cyclone energy from the re-forecasts to allow easy comparison with model climate data and to determine the significance level of the results. Links: FUG | |
Hovmoeller diagram time-longitude diagram | Hovmoeller or time-longitudes diagrams show the time evolution of the ensemble mean anomaly of a parameter (previous or observed data above the horizontal line, and forecast data below the horizontal line). It highlights the behaviour and progression of atmospheric features, particularly in the tropics. Links: FUG | |
Terms associated with satellite application facilities | ||
H SAF | Hydrological satellite application facility. Links: Eumetsat | |
ROM SAF | Radio occultation meteorology satellite application facility. Links: Eumetsat | |
O3M SAF | Ozone and atmospheric chemistry monitoring satellite application facility. Links: Eumetsat | |
LSA SAF | Land surface analysis satellite application facility. Links: Eumetsat | |
NWP SAF | Numerical weather prediction satellite application facility. Links: Eumetsat | |
OSI SAF | Ocean and sea-ice satellite application facility. Links: Eumetsat | |
Terms associated with satellite observations | ||
ATOVS | Advanced TIROS operational vertical sounder. Links: Eumetsat | |
AVHRR | Advanced very high resolution radiometer. Links: Eumetsat | |
AAPP | ATOVS and AVHRR pre-processing package. Links: Eumetsat | |
MODIS | Moderate resolution imaging spectroradiometerspectro-radiometer. Links: NASA | |
CRYOSAT-2 | Sea ice mission orbiting satellite. Links: ESA | |
GPS-RO | Lim sounding of boundary layer. Links: ECMWF | |
IMS | NOAA/NESDIS interactive multi-sensor snow and ice mapping system. Links: NOAA | |
SMOS ASCAT | Soil moisture and ocean salinity satellite mission. Satellite system measuring surface emission that is strongly related to soil moisture over continental surfaces, salinity and surface state of the oceans, and to the thickness of sea ice. Links: ECMWF | |
TERRA, AQUA | Sun-synchronous orbital satellites with high resolution energy monitoring and imaging systems. Links: TERRA, AQUA | |
scatterometer | Satellite-borne microwave radar sensors that give a measure of the roughness of the ocean surface and allow derivation of wave height and surface winds. Links: Eumetsat | |
...