Last modified on Oct 09, 2023 16:09

Contributors:

1. History of modifications to this Product User Guide

2. Acronyms

3. Introduction

The Copernicus Climate Change Service (C3S) aims to provide authoritative information about the past, present and future climate in Europe and the rest of the World. It is in this context that the pan-European (Fig. 1) Copernicus European Regional ReAnalysis (CERRA) has been produced, under the contract C3S_322_Lot1. The land surface reanalysis dataset, CERRA-Land, spans the period September 1984 to April 2021 and has been produced at a horizontal spatial resolution of 5.5km. The dataset can be used in support of adaptation action and policy development as well as contribute to climate monitoring and research, but also provide valuable information to climate services.

Figure 1: CERRA-Land domain. Orography [m] is presented in color. The system used the Lambert Conformal Conic projection.

The need for precipitation and surface variables at an ever-increasing spatial and temporal resolution is a recurrent demand. These variables allow, among other things, to address water resource management issues and to carry out climate change impact studies. Regional surface reanalyses are a way to reconstruct these variables for past periods covering several decades using state-of-the-art models.

A common way to generate a land surface reanalysis dataset is to run in an open loop a land surface model driven by gridded atmospheric datasets (forcing data) to describe the evolution of water and energy cycles over land. The model is run offline, i.e. without feedback on the atmospheric analysis from the assimilation cycle, to prevent propagation and increase of errors due to surface-atmosphere coupling. The surface observations were only used for the construction of the forcing data. This offline system allows consistency with mass conservation which is essential for conducting climate change impact studies. The quality of the reanalysis depends on both the quality of the forcing data and the parameterization of the surface and ground physical processes. This approach has been used by the ERA5-Land, MERRA-Land, ERA-Interim-Land global land surface reanalysis  (Albergel et al., 2013; Muñoz Sabater et al., 2021), but also by regional reanalysis such as MESCAN-SURFEX (Bazile et al., 2017). The new CERRA-Land dataset follows the same approach.

 CERRA-Land is the result of a unique standalone integration of the SURFEX V8.1 land surface model (Le Moigne et al. 2020) driven by meteorological forcing data from the CERRA atmospheric reanalysis and an offline analysis of daily, 24-h, accumulated surface precipitation (Soci et al., 2016). The accumulated precipitation analysis is created by an optimal interpolation between an initial estimate (first guess) based on CERRA predicted precipitation and daily rain gauge data. Temperature and relative humidity forcing data at 2 metres are from the CERRA surface analysis and the short- and long-wave downwelling radiation, wind speed at 10 metres, and surface pressure are from CERRA forecast data. A brief description of the CERRA-Land system is available in Documentation tab.

4. Details about the data fields

The outputs of the land surface model were archived as forecast type at hourly time step and the precipitation analysis was archived at daily time step (as analysis type). The ouputs data are available in the same way as the CERRA dataset as if analysis were done every 3 hours ( at 00, 03, 06, 09, 12, 15, 18 and 21 UTC). The date of the analysis is related to the two metre temperature and humidity forcing fields. The forecast fields are available at three time step ( +1, 2  and 3 hours).

The CERRA-Land system uses the tiling approach where each grid box (5.5 km X 5.5 km) of the model is divided into three different fractions: urban, lake and natural land.

The surface–atmosphere fluxes are then aggregated for each atmospheric grid box, according to the fraction (frac) of the three types of surface in the cell. The averaged value (F) over the grid cell is thus given by

where the values F_nature, F_urban, and F_lake are calculated by specific physical parametrization. For costal grid point, the positive values of sea fraction less than 1 were set to 0 and frac_nature was set to 1 - (frac_urban + frac_lake). The sum of fractions must always equal 1. It means that the costal aggregated values can be biased because no sea fraction was considered.

The surface pressure, 10-metre wind speed and direction, two metre temperature and relative humidity analysis were not archived into the CERRA-Land dataset, but are available through the CDS from the CERRA on single levels dataset.

The available CERRA-Land variables in the CDS are either instantaneous, accumulated or static. The natural, urban and inland water fraction will be available in netCDF format only. This is specified for each of the variables listed in the tables below.

4.1. Overview of variables calculated as a mean value of a grid box

Table 1: Overview of variables calculated as a mean value of a grid box.

4.1.1. Albedo

The albedo [0-100%] is the total reflectance of downward solar radiation at the surface over the grid box. The albedo is the ratio of one hour time-integrated surface solar radiation upward by the one hour time-integrated surface solar radiation downward. Multiplying the albedo with the one hour accumulated downward solar radiation gives the one hour accumulated upward solar radiation.

4.1.2. Evaporation

Evaporation is the amount of water that has evaporated from the earth’s surface from the initial time of the forecast to the forecast time step. It is given as a mean for the grid area between the three surface types in the grid - inland water, natural land and urban. Hence, evaporation is represented by negative values and positive values represent condensation. By model convention downward fluxes are positive. It is an accumulated variable.

4.1.3. Total precipitation

The total precipitation is the amount of precipitation falling at the surface during the last 24-hours. It includes all kind of precipitation forms as convective precipitation, large scale precipitation, liquid and solid precipitation. The total precipitation is available only for the analyses at 06 UTC. It is an accumulated field from the previous day at 06 UTC to the present day at 06 UTC. The date in the metadata GRIB2 file represents the end of the accumulation period.

4.1.4. Skin temperature 

It represents the average air temperature at the surface of each grid box. The skin temperature is an average of temperatures given by the three surface types in the grid - inland water, natural land and urban. It is an instantaneous variable.

4.1.5. Surface latent heat flux

The surface latent heat flux is the accumulated exchange of latent heat (due to phase transitions - evaporation, condensation) with the surface (ground and water) through turbulent diffusion from the initial time of the forecast to the forecast time step. It is given as a mean for the grid area between the three surface types in the grid - inland water, natural land and urban. By model convention downward fluxes are positive. It is an accumulated variable.

4.1.6. Surface sensible heat flux

The surface sensible heat flux is the accumulated exchange of heat (no phase transition) with the surface (ground and water) through turbulent diffusion from the initial time of the forecast to the forecast time step. It is given as a mean for the grid area between the three surface types in the grid - inland water, natural land and urban. By model convention downward fluxes are positive. It is an accumulated variable.

4.1.7. Surface net solar radiation

The surface net solar radiation is the accumulated solar short-wave radiation that is absorbed at the surface from the initial time of the forecast to the forecast time step. It is calculated as the difference between the downward solar energy and the upward solar energy at the surface. By model convention downward fluxes are positive. It is an accumulated variable.

4.1.8. Surface solar radiation downwards

The surface solar radiation downward is the accumulated total (direct and diffuse) solar short-wave radiation reaching the surface from the initial time of the forecast to the forecast time step. By model convention downward fluxes are positive. It is an accumulated variable.

4.1.9. Surface net thermal radiation

The net thermal radiation at the surface is accumulated from the initial time of the forecast to the forecast time step. It is calculated as the difference between the thermal radiation downwards and the upward thermal radiation at the surface. By model convention downward fluxes are positive. It is an accumulated variable.

4.1.10. Surface thermal radiation downwards

The surface thermal radiation downward is the amount of thermal (long-wave) radiation reaching the surface accumulated from the initial time of the forecast to the forecast time step. By model convention downward fluxes are positive. It is an accumulated variable.

4.1.11. Soil heat flux 

The soil heat flux is the energy receive by the soil to heat it per unit of surface and time. The soil heat flux is positive when the soil receives energy (warms) and negative when the soil loses energy (cools). It is an instantaneous variable.

4.1.12. Surface roughness 

The surface roughness describes the aerodynamic roughness length. It is given as a mean for the grid area between the three surface types in the grid (inland water, natural land and urban) and has missing values over the ocean. The roughness length of the surface is the height above the surface at which the wind profile is assumed to become zero. Each grid point has one value representing the mean over the grid point. The effective surface roughness is depending on the orographic component (constant part), the snow depth, the evolution of the Leaf Area Index and the fraction of vegetation, which is different for each month. It is an instantaneous variable.

4.2. Overview of variables available for the natural land fraction

4.2.1. snow variables

The snow variables are related to natural land only (e.g the fraction of snow cover which represents the fraction of natural land which has snow on the ground). Snow on urban or lake fraction is not available.

Table 2: Overview of variables available for the natural land fraction.

4.2.1.1.  Fraction of snow cover

It represents the fraction of natural land covered by snow. It is an instantaneous variable and takes values between 0 and 1.

4.2.1.2.  Snow depth

Snow thickness on the ground. It is an instantaneous variable.

4.2.1.3.  Snow depth water equivalent

The mass of liquid water obtained from melting the snow per unit area. This is equivalent to the depth of the liquid water in units of mm. It is an instantaneous variable.

4.2.1.4.  Snow density

The mean snow density is calculated as the ratio of snow depth water equivalent by the snow depth. It is an instantaneous variable.

4.2.1.5. Snow albedo

It is defined as the fraction of solar (short-wave) radiation reflected by the snow, across the solar spectrum, for both direct and diffuse radiation. Values vary between 0 and 1. It is an instantaneous variable.

4.2.1.6.  Temperature of snow layer

It is the mean temperature of the 12 snow layers. It is an instantaneous variable.

4.2.1.7.  Snow melt

Melting of snow. This variable is accumulated from the beginning of the forecast time to the end of the forecast step. It is an accumulated variable.

4.2.2. Others surfaces variables

4.2.2.1. Percolation

The mass per unit area of water that drains below the deepest soil level in the model. The percolation (or drainage) is accumulated from the initial time of the forecast to the forecast time step. This variable is calculated for the natural land, including soil, vegetation and snow (not for urban and water bodies fraction). It is an accumulated variable calculated for the natural land, including soil, vegetation and snow whereas.

4.2.2.2. Surface runoff

The mass per unit area of water at the surface when saturation occurs.  It is an accumulated variable calculated for the natural land, including soil, vegetation and snow.

4.2.3. Soil variables

The prognostic variables of soil temperature and soil moisture are represented in the model by a diffusive approach. Such a method proposes a discretisation of the soil into 14 layers, resulting in a total depth of 12 m, with a fine description of the subsurface layers to capture the diurnal cycle. The vertical discretisation (bottom depth of each layer in metres) is as follows: 0.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, and 12 m. Heat transfer is resolved over the total depth, while moisture transfer is resolved only over the depth of the roots, which depends on the type of vegetation and its geographical location.

4.2.3.1. Soil temperature

The soil temperature is provided for each soil layer. The SURFEX model has 14 soil layers. The vertical discretisation (bottom depth of each layer in metres) is as follows: 0.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, and 12 m. Soil temperature is an instantaneous variable.

4.2.3.2. Volumetric soil moisture 

The volume concentration of liquid and ice water. The vertical discretisation (bottom depth of each layer in metres) is as follows: 0.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, and 12 m. The volumetric soil moisture is available for each soil layer. It is an instantaneous variable.

4.2.3.3. Liquid volumetric soil moisture (non-frozen)

It is the volume concentration of liquid water only. The vertical discretisation (bottom depth of each layer in metres) is as follows: 0.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, and 12 m. A liquid volumetric soil moisture is available for each soil layer. It is an instantaneous variable.

4.3. Overview of variables available for the inland water fraction.

Table 3: Overview of variables available for the inland water fraction.

4.3.1.1. Lake bottom temperature

Temperature of water at the bottom of inland water bodies (lakes). The model keeps lake depth and surface area (or fractional cover) constant in time. It is an instantaneous variable.

4.3.1.2. Lake ice depth

The thickness of ice on inland water bodies (lakes). A single ice layer is represented. This parameter is the thickness of that ice layer. It is an instantaneous variable.

4.3.1.3. Lake ice temperature

The temperature of the uppermost surface of ice on inland water bodies (lakes). A single ice layer is represented. It is an instantaneous variable.

4.3.1.4. Lake mix-layer depth

The thickness of the upper most layer of an inland water body (lake) that is well mixed and has a near constant temperature with depth (uniform distribution of temperature). The Flake model  represents inland water bodies with two layers in the vertical, the mixed layer above and the thermocline below. Thermoclines upper boundary is located at the mixed layer bottom, and the lower boundary at the lake bottom. Mixing within the mixed layer can occur when the density of the surface (and near-surface) water is greater than that of the water below. Mixing can also occur through the action of wind on the surface of the lake. It is an instantaneous variable.

4.3.1.5. Lake mix-layer temperature

The temperature of the upper most layer of inland water bodies (lakes) that is well mixed. The Flake model represents inland water bodies with two layers in the vertical, the mixed layer above and the thermocline below. Thermoclines upper boundary is located at the mixed layer bottom, and the lower boundary at the lake bottom. Mixing within the mixed layer can occur when the density of the surface (and near-surface) water is greater than that of the water below. Mixing can also occur through the action of wind on the surface of the lake. It is an instantaneous variable.

4.3.1.6. Lake shape factor

This parameter describes the way that temperature changes with depth in the thermocline layer of inland water bodies (lakes). It is used to calculate the lake bottom temperature and other lake-related parameters. The Flake model represents inland water bodies with two layers in the vertical, the mixed layer above and the thermocline below where temperature changes with depth. It is an instantaneous variable.

4.3.1.7. Lake total layer temperature

The mean temperature of total water column in inland water bodies (lakes). The Flake model represents inland water bodies with two layers in the vertical, the mixed layer above and the thermocline below where temperature changes with depth. This parameter is the mean over the two layers. It is an instantaneous variable.

4.4. Static variables

Static variable do not change depending on the model initial time or the forecast length (in other words they are time-independent). These include the land-sea mask, that is the fraction of land in a given model grid box of 5.5 x 5.5 km² in units of %, and the orography in units of m. For each model grid box in CERRA-Land, 3 tile fractions are defined in units of %: (1) the fraction of inland water (lakes and rivers), (2) the fraction of urban areas, and (3) the fraction of nature, i.e. land areas that are not inland water or urban. There is no data available for the fraction of sea. The fraction data will be available directly from the climate data store website in netCDF format.

Table4: Static variable. The variables marked with * and labelled with TBD ("To Be Determined") do not have yet short

names in the WMO GRIB2 code definitions and have not been uploaded to the CDS in the first batch of released data.

The  fraction data can be downloaded as netcdf file format: CERRALand_tiles_fraction.nc

4.4.1.1. Lake depth

Depth of inland water. It is defined for positive fractions only. It is a static variable.

4.4.1.2. Volumetric wilting point

Model soil water content at which the vegetation wilts and can no longer recover. When the soil moisture reaches the wilting point, the vegetation is not able to extract the soil water. The soil moisture content is too low to be absorbed by the vegetation. The vertical discretisation (bottom depth of each layer in metres) is as follows - 0.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, and 12 m. It is a static variable.

4.4.1.3. Volumetric transpiration stress-onset (soil moisture)

The soil moisture is the water content of a soil after gravitational drainage. When the water content of the soil reaches this value, the water cannot drain any more by gravity. The vertical discretisation (bottom depth of each layer in metres) is as follows - 0.01, 0.04, 0.1, 0.2, 0.4, 0.6, 0.8, 1, 1.5, 2, 3, 5, 8, and 12 m. It is a static variable.

4.4.1.4. Land-sea mask

The land-sea mask is a field that contains, for every grid point, the proportion of land (including inland water) in the grid box. It is the sum of the three fractions - natural land, urban and inland water. The variable is dimensionless and takes values between 0 (sea) and 1 (land). It is a static variable.

4.4.1.5. Orography

The height above sea level of the land surface. This variable does not change with snow cover. It is a static variable.

4.4.1.6. Inland water tile fraction

This variable represents the fraction of water (e.g lakes, rivers) in the grid-box. It takes values between 0 and 1. It is a static variable.

4.4.1.7. Urban tile fraction

This variable represents the urban (e.g town) in the grid box. It takes values between 0 and 1. It is a static variable.

4.4.1.8. Nature tile fraction

This variable represents the fraction of natural land areas that are neither inland water nor urban in the grid box. It takes values between 0 and 1. It is a static variable.

4.5. Accumulated surface fluxes

All energy fluxes at the surface are accumulated variables from the initial time of the forecast to the forecast hour in question with the unit J/m². They are considered positive downward to the surface. Average hourly energy fluxes in W/m² can be computed by subtracting two successive hourly accumulated variables and dividing by 3600 s. The albedo in units of % is defined by the ratio of average hourly upward radiation upward by the average hourly solar radiation downward. The net solar radiation is the difference between the solar radiation downward and the solar radiation upward. The solar radiation upward can be calculated by subtracting the net solar radiation from the downward solar radiation.

The net thermal radiation is the difference between the downward thermal radiation and the upward thermal radiation. The upward thermal radiation can be calculated by subtracting the net thermal radiation from the downward thermal radiation. The accumulated surface sensible heat flux is the conductive energy from the atmosphere to the surface. If this is going from the surface to the atmosphere it has negative values. The accumulated latent heat flux is the sum of all latent energy fluxes that are due to the phase transitions of water. Here condensation causes a positive latent heat flux to the surface and evaporation causes a negative heat flux from the surface.

5. References

Albergel C., Dorigo W., Reichle R. H., Balsamo G., de Rosnay P., Muñoz-Sabater J., Isaksen L., de Jeu R., and Wagner W.: Skill and Global Trend Analysis of Soil Moisture from Reanalyses and Microwave Remote Sensing, Journal of Hydrometeorology, 14, 1259–1277, 825, https://doi.org/10.1175/JHM-D-12-0161.1, 2013

Bazile E., Abida R., Verelle A., Le Moigne P., Szczypta C., (2017): MESCAN-SURFEX surface analysis, deliverable D2.8 of the UERRA project, http://www.uerra.eu/publications/deliverable-reports.html

Le Moigne P., Bazile E., Glinton M. and Verrelle A. (2022) : Documentation of the CERRA-Land system ( C3S delivrable C3S_D322_Lot1.1.1.12_202110_documentation_CERRA-Land)

Le Moigne P., Besson F., Martin E., Boé J., Boone A., Decharme B., Etchevers P., Faroux S., Habets F., Lafaysse M., Leroux D., and Rousset-Regimbeau F.: The latest improvements with SURFEX v8.0 of the Safran–Isba–Modcou hydrometeorological model for France, Geosci. Model Dev., 13, 3925–3946, https://doi.org/10.5194/gmd-13-3925-2020, 2020.

Muñoz-Sabater J., Dutra E., Agustí-Panareda A., Albergel C., Arduini G., Balsamo G., Boussetta S., Choulga M., Harrigan S., Hersbach H., Martens B., Miralles D. G., Piles M., Rodríguez-Fernández N. J., Zsoter E., Buontempo C., and Thépaut J.-N.: ERA5-Land: A state-of-the-art global reanalysis dataset for land applications, Earth Syst. Sci. Data,13, 4349–4383, 2021. https://doi.org/10.5194/essd-13-4349-2021 

Schimanke S., Isaksson L., and  Edvinsson L.: CERRA data user guide (C3S deliverable C3S_322_Lot1.4.1.3_CERRA_ data_user_guide)

Soci C., Bazile E., Besson F., and Landelius T. (2016). High-resolution precipitation re-analysis system for climatological purposes. Tellus A, Dynamic Meteorology and Oceanography, 68:1, DOI: 10.3402/tellusa.v68.29879

6. Related articles

• Page:
  CERRA-Land surface reanalysis: Data User Guide
• Page:
  Copernicus European Regional ReAnalysis (CERRA): product user guide
• Page:
  Copernicus European Regional ReAnalysis (CERRA)