Contributors: E. Carboni (UKRI-STFC RAL Space), G.E. Thomas (UKRI-STFC RAL Space)
Issued by: STFC RAL Space (UKRI-STFC) / Elisa Carboni
Date: 22/07/2022
Ref: C3S2_D312a_Lot1.1.3.1-v4.0_202207_PQAD_CCISurfaceRadiationBudget_v1.0
Official reference number service contract: 2021/C3S2_312a_Lot1_DWD/SC1
History of modifications
List of datasets covered by this document
Related documents
Acronyms
List of tables
List of figures
General definitions
The “CCI product family” Climate Data Record (CDR) consists of two parts. The ATSR2-AATSR Surface Radiation Budget CDR is formed by a TCDR brokered from the ESA Cloud_cci project and an ICDR derived from the SLSTR on board of Sentinel-3. ICDR uses the same processing and infrastructure as the TCDR. Both TCDR and ICDR data have been produced by STFC RAL space.
These Surface Radiation Budget datasets from polar orbiting satellites consist of seven main variables: Surface Incoming Shortwave radiation (SIS), Surface Reflected Shortwave radiation (SRS), the Surface Net Shortwave radiation (SNS), the Surface Outgoing Longwave radiation (SOL), Surface Downwelling Longwave radiation (SDL), Surface Net Longwave radiation (SNL), and the Surface Radiation Budget (SRB).
Bias (accuracy): Mean difference between TCDR/ICDR and reference data
\( b=\frac{\sum_{i=1}^N (p_i - r_i)}{N} \ \ (Eq. 1) \)
Where: pi is the CDR product, b is the mean bias and ri is the equivalent value from the reference dataset. N is the number of observations.
bc-RMSE (precision): Bias corrected root mean squared error to express the precision of TCDR/ICDR compared to a reference data record
\( bc- RMSE=\sqrt{\frac{\sum_{i=1}^N ((p-b)-r)^2}{N}} \ \ (Eq. 2) \)
Where: pi is the CDR product, b is the mean bias and ri is the equivalent value from the reference dataset. N is the number of observations.
Stability: The variation of the bias over a multi-annual time period
Table 1: Summary of variables and definitions
Variables | Abbreviation | Definition |
Surface incoming solar radiation | SIS | The total incoming solar flux, measured at the Earth’s surface. |
Surface reflected solar radiation | SRS | The total upwelling shortwave flux, measured at the Earth’s surface. |
Surface net solar radiation | SNS | The net downwelling solar flux, measured at the surface (equal to SIS – SRS). |
Surface downwelling longwave radiation | SDL
| The total downwelling thermal infrared flux, measured at the Earth’s surface. |
Surface outgoing longwave radiation
| SOL
| The total upwelling thermal infrared flux, measured at the Earth’s surface. |
Surface net longwave radiation | SNL | The net downwelling thermal infrared flux, measured at the Earth’s surface (equal to SDL-SOL). |
Total surface radiation budget | SRB | The total net downwelling radiative flux, measured at the Earth’s surface (equal to (SIS+SDL) – (SRS+SOL)). |
Table 2: Definition of processing levels
Processing level | Definition |
Level-1b | The full-resolution geolocated radiometric measurements (for each view and each channel), rebinned onto a regular spatial grid. |
Level-2 (L2) | Retrieved cloud variables at full input data resolution, thus with the same resolution and location as the sensor measurements (Level-1b). |
Level-3C (L3C) | Cloud properties of Level-2 orbits of one single sensor combined (averaged) on a global spatial grid. Both daily and monthly products provided through C3S are Level-3C. |
Table 3: Definition of various technical terms used in the document
Jargon | Definition |
Brokered product | The C3S Climate Data Store (CDS) provides both data produced specifically for C3S and so-called brokered products. The latter are existing products produced under an independent programme or project which are made available through the CDS. |
Climate Data Store (CDS) | The front-end and delivery mechanism for data made available through C3S. |
Retrieval | A numerical data analysis scheme which uses some form of mathematical inversion to derive physical properties from some form of measurement. In this case, the derivation of cloud properties from satellite measured radiances. |
Forward model | A deterministic model which predicts the measurements made of a system, given its physical properties. The forward model is the function which is mathematically inverted by a retrieval scheme. In this case, the forward model predicts the radiances measured by a satellite instrument as a function of atmospheric and surface state, and cloud properties. |
TCDR | It is a consistently-processed time series of a geophysical variable of sufficient length and quality. |
ICDR | An Interim Climate Data Record (ICDR) denotes an extension of TCDR, processed with a processing system as consistent as possible to the generation of TCDR. |
CDR | A Climate Data Record (CDR) is defined as a time series of measurements with sufficient length, consistency, and continuity to determine climate variability and change. |
Scope of the document
This document provides a description of the product validation methodology for the Essential Climate Variable (ECV) Surface Radiation Budget. This CDR comprises inputs from two sources: (i) brokered products from the Cloud Climate Change Initiative (ESA’s Cloud_cci), namely those coming from processing of the Advanced Along-Track Scanning Radiometer (A)ATSR) data and (ii) those produced under this contract fore, specifically those coming from processing of the Sea and Land Surface Temperature Radiometers (SLSTR).
The Thematic Climate Data Record (TCDR) is the product brokered from the European Space Agency Cloud Climate Change Initiative (ESA’s Cloud_cci) ATSR2-AATSR version 3.0 (Level-3C) dataset. This is produced by STFC RAL Space from the second Along-Track Scanning Radiometer (ATSR-2) on board the second European Remote Sensing Satellite (ERS-2) spanning the period 1995-2003, the Advanced ATSR (AATSR) on board ENVISAT spanning the period 2002-2012.
In addition, the Interim Climate Data Record (ICDR) is the product derived from the SLSTR on board of Sentinel-3 and spans the period from 2017 to present.
Validation of ATSR2, AATSR and SLSTR derived products for the period from January 2017 to December 2021 are described in this document. It summarizes and refers to the methodology presented in the Cloud_cci Product Validation and Intercomparison Report [D1], used in the validation of the TCDR product. The same methodology is applied to the ICDR dataset.
Executive Summary
The ESA Climate Change Initiative (CCI) Surface Radiation Budget Climate Data Record (CDR) is a brokered product from the ESA Cloud_cci project, while the extension Interim CDR (ICDR) produced from the Sea and Land Surface Temperature Radiometer (SLSTR) is produced specifically for C3S. The product is generated by STFC RAL Space, using the Community Cloud for Climate (CC4CL) processor, based on the Optimal Retrieval of Aerosol and Cloud (ORAC) algorithm. The Surface Radiation Budget is a product of the Broadband Radiative Flux Retrieval (BRFR) module of CC4CL, which uses the cloud properties produced by ORAC to compute broadband radiative flux values.
The Cloud_cci record comprises 17 years (1995-2012) of satellite-based measurements derived from the Along Track Scanning Radiometers (ATSR-2 and AATSR) onboard the ESA second European Research Satellite (ERS-2) and ENVISAT satellites. This CDR is partnered with the ICDR produced from the Sentinel-3A SLSTR, beginning in 2017, and Sentinel-3B SLSTR beginning in October 2018. In addition to individual products from each Sentinel-3 platform, a combined product that averages data from both SLSTR instruments into single daily and monthly means will also be provided.
The dataset encompasses level-3 data (monthly means) on a regular global latitude-longitude grid (with a resolution of 0.5°´ 0.5°) and includes these products: the Surface Incoming and Reflected Shortwave radiation (SIS and SRS respectively), the Surface Downwelling and Outgoing Longwave radiation (SDL and SOL respectively), the Surface Net Shortwave and Longwave radiation (SNS and SNL), and the total Surface Radiation Budget (SRB).
This document is divided into different sections:
- the first section presents a brief description of the surface radiation CDR products together with reference for further information;
- the second section presents the datasets used to estimate the accuracy of the CDR surface radiation dataset;
- the third section presents the methodology used for the validation and is divided in different subsections that describe: the validation with ground measurements, the comparison with Clouds and Earth Radiation Energy System (CERES) surface data and the uncertainty propagation used to estimate the accuracy of the other parameters.
1. Validated products
The ATSR2-AATSR Surface Radiation Budget CDR is formed by a TCDR brokered from the ESA Cloud_cci project and an ICDR derived from the SLSTR on board of Sentinel-3. Both TCDR and ICDR data have been produced by STFC RAL space.
The SLSTR ICDR, both from the individual instruments (version 3.0) and combining both in a single product (version 4.0), is supplied to the CDS via the same route and uses the same processing software and infrastructure as the TCDR. The retrieval algorithm is described in detail in [D2].
These Surface Radiation Budget datasets from polar orbiting satellites consist of: Surface Incoming Shortwave radiation (SIS), Surface Reflected Shortwave radiation (SRS), the Surface Net Shortwave radiation (SNS), the Surface Outgoing Longwave radiation (SOL), Surface Downwelling Longwave radiation (SDL), Surface Net Longwave radiation (SNL), and the Surface Radiation Budget (SRB).
The datasets cover the period from June 1995 to April 2012 (TCDR), using satellite-based measurements derived from ATSR2 and AATSR onboard the polar orbiting ERS-2 and ENVISAT respectively, and the period from January 2017 onwards using the SLSTR measurements (ICDR). These are level 3 products (monthly means) on a regular global latitude-longitude grid (with 0.5° x 0.5° resolution). Cloud properties from the ESA Cloud_cci dataset version 3 (TCDR) are used for the estimation of the Surface Radiation Budget1. The Cloud_cci dataset can be downloaded here: https://climate.esa.int/en/projects/cloud/data/. The SLSTR based ICDR extends the coverage, with a five year gap, from 2017 onwards and is only available through the Copernicus Climate Data Store (CDS). Table 1-1 reports the values from the PQAR[D4]
The TCDR dataset that includes Surface Radiation Budget products as well as Cloud Properties and Earth Radiation Budget products are described by Poulsen et al. (2019) [D3].
Table 1-1: Summary of the accuracy of the Surface Radiation Budget dataset (taken from [D4]). The accuracies presented in ´bold´ come from direct validation with a ground measurement network, the other come from the intercomparison with similar datasets or with an uncertainty propagation. ICDR values are obtained from data between January 2017 and December 2021 for SLSTR-A and between October 2018 and December 2021 for SLSTR-B.
Product name | TCDR Accuracy [W/m2] | ICDR SLSTR-A Accuracy [W/m2] | ICDR SLSTR-B Accuracy [W/m2] | |
Surface Incoming Shortwave radiation (SIS) | 8.2 | 0.5 | 1.5 | |
Surface Reflected Shortwave radiation (SRS) | 4.6 | 1.8 | 2.0 | |
Surface Net Shortwave radiation (SNS) | 13 | 2.3 | 3.5 | |
Surface Outgoing Longwave radiation (SOL) | 11 | 1.5 | 3.9 | |
Surface Downwelling Longwave radiation (SDL) | 12 | 13 | 12 | |
Surface Net Longwave radiation (SNL) | 23 | 15 | 16 | |
Surface Radiation Budget (SRB) | 36 | 17 | 20 |
2. Description of validating datasets
The Surface Radiation Budget TCDR dataset from the ATSR2 and AATSR instruments is compared against the ground measurements dataset: the central archive of the Baseline Surface Radiation Network (BSRN)2.
BSRN stations measure direct, diffuse and global downwelling shortwave and longwave fluxes in 1 min temporal resolution. The manned stations are located at locations, which are representative of a relatively large surrounding area for the use in satellite and climate model validation. The quality controlled datasets are available for the years 1992 to 2017 in ASCII file format. Specially calculated monthly means of daily mean products have been used in TCDR validation [D1]
Both, TCDR and ICDR from SLSTR instruments, are compared with the Clouds and Earth Radiation Energy System (CERES) Energy Balanced and Filled (EBAF) fluxes Edition 4.1 Top of atmosphere (TOA) and Bottom of Atmosphere (BOA) fluxes Edition (Loeb et al., 2018)3.
The CERES product provides long-term shortwave (SW) and longwave (LW) TOA fluxes for all- and clear-sky conditions. The CERES instruments fly on the Terra and Aqua satellites and cover a period from March 2000 to June 2002 for Terra only, and cover combined Terra and Aqua observations from July 2002 to January 2017. The CERES instruments provide global coverage daily, and monthly mean regional fluxes and are based upon daily samples over the entire globe.
In addition to the TOA fluxes the CERES dataset provides EBAF Ed4.0 Surface Fluxes. EBAF Surface fluxes (used to compare with the CDR dataset) are derived using CERES TOA products and coincident imager data from the Moderate Resolution Imaging Spectrometer (MODIS) and the Visible Infrared Imaging Radiometer Suite (VIIRS).
3. Description of product validation methodology
The validation strategy is described in section 2.4 of [D1].
The methodology uses the bias between the Cloud_cci product and the reference data to estimate the accuracy of the dataset.
The bias corrected root mean squared error (bc-RMSE) is used to express the precision of CDR compared to a reference data record, which is also known as the standard deviation from the mean.
The SIS and SDL products of the TCDR dataset are validated against ground measurements and compared with the CERES satellite dataset in [D1].
The accuracy for SIS and SDL (TCDR) are estimated using the ground measurements as reference because these are considered to be more accurate than satellite measurements.
The SIS and SDL products of the ICDR dataset are evaluated by a comparison with the CERES dataset and the evaluation is performed within the C3S project.
The SRS, SNS SNL and SRB accuracies for both TCDR and ICDR are estimated by uncertainty propagation as explained below (section 3.3 to 3.5). Table 3-1 summarizes the methodology used to estimate the accuracies for each product.
In all cases, the same validation approach will be applied to the combined SLSTR product (version 4.0) as is used for the individual platform SLSTR data (version 3.0 and 3.1).
Table 3-1: Summary of methodologies used to estimate the accuracies, for TCDR and ICDR datasets
Product name | Validation with BSRN | Comparison with CERES | Uncertainty propagation |
Surface Incoming Shortwave radiation (SIS) | TCDR | TCDR and ICDR | |
Surface Reflected Shortwave radiation (SRS) | ICDR | TCDR and ICDR | |
Surface Net Shortwave radiation (SNS) | TCDR and ICDR | ||
Surface Outgoing Longwave radiation (SOL) | TCDR and ICDR | ||
Surface Downwelling Longwave radiation (SDL) | TCDR | TCDR and ICDR | |
Surface Net Longwave radiation (SNL) | TCDR and ICDR | ||
Surface Radiation Budget (SRB) | TCDR and ICDR |
3.1 Validation with BSRN ground base radiative flux
BSRN stations measure direct, diffuse and global downwelling shortwave and longwave fluxes in 1 min temporal resolution. The 1-minute data were aggregated to monthly averages which were used as validation data. Using the TCDR and the reference datasets (in different locations around the world) we compute the bias and standard deviation.
The validation method for Surface Incoming Shortwave radiation (SIS) and Surface Downwelling Longwave radiation (SDL) with BSRN ground measurements is described in sections 2.4 and 3.3.2 of [D1].
3.2 Comparison with CERES satellite data
TCDR and reference datasets are compared by calculation of multi-annual mean (i.e., we produce a global map of one parameter averaged over multiple years, we calculate the mean of this global map and compare it with the equivalent mean from reference data) and standard deviation for the common time period (2003-2011). For the ICDR, we used the CERES dataset as a reference and compared the means of multi-monthly means (i.e., we compute the mean of the differences between CDR monthly mean global averages and reference data monthly mean global averages) as well as the standard deviation for the time period 2017-01 to 2021-12. Global maps of multiannual Surface Incoming Shortwave radiation (SIS) and Surface Downwelling Longwave radiation (SDL) are computed for the CDR and the reference dataset. The scores (bias and bc-RMSE) are calculated by including all valid data points pairwise in the CERES dataset and the CDR. The same methodology is applied for the TCDR in [D1] and ICDR within C3S.
The validation method for Surface Incoming Shortwave radiation (SIS) and Surface Downwelling Longwave radiation (SDL) with CERES is described in section 5.3 and 5.4 of [D1] as well as the results of the accuracy (bias) ΔSIS/ΔSDL. The same methodology will be used to estimate the accuracy of SOL in comparison with CERES.
3.3 Surface Reflected Shortwave Radiation (SRS)
The accuracy of SRS is estimated from the accuracy of Surface Incoming Shortwave radiation (SIS) and Surface albedo (SAL). Applying the error propagation, the accuracy of the SRS product can be estimated as:
\( \Delta SRS= \frac{\delta SRS}{\delta SIS} \Delta SIS + \frac{\delta SRS}{\delta SAL} \Delta SAL = SAL \Delta SIS + SIS \Delta SAL, \quad \ \ (Eq. 3) \)
Where
\( \Delta SIS \)
comes from [D1] and
\( \Delta SAL \)
is considered as 25% of the SAL value. SAL is estimated as the ratio between Surface Reflected Shortwave radiation (SRS) and Surface Incoming Shortwave radiation (SIS).
\( SAL = SRS / SIS, \quad \ \ (Eq. 4) \)
3.4 Surface Net Shortwave Radiation (SNS)
The Surface Net Shortwave radiation (SNS) is calculated using:
\( SNS = SIS - SRS, \quad \ \ (Eq. 5) \)
And the accuracy will be estimated as:
\( \Delta SNS = \Delta SIS + \Delta SRS, \quad \ \ (Eq. 6) \)
3.5 Surface Net Longwave Radiation (SNL)
Surface Net Longwave radiation (SNL) is calculated [D2] from:
\( SNL = SDL - SOL, \quad \ \ (Eq. 7) \)
The accuracy
\( \Delta SNL \)
will be estimated as:
\( \Delta SNL = \Delta SDL + \Delta SOL, \quad \ \ (Eq. 8) \)
3.6 Surface Radiation Budget (SRB)
The total Surface Radiation Budget (SRB) is simply the sum of the short and longwave contributions:
\( SRB = SNS + SNL, \quad \ \ (Eq. 9) \)
The accuracy will be estimated as:
\( \Delta SRB = \Delta SNS + \Delta SNL, \quad \ \ (Eq. 10) \)
4. Summary of validation results
The TCDR validation results are provided in [D1], section 3.3.2, 5.3 and 5.4.
As an example Figure 4-1 from [D1] shows the results of the TCDR comparison with the BSRN incoming shortwave (SIS) and longwave (SDL) radiation with scatter plots and global maps showing the bias for each station. A more detailed description and analysis of the results is available in the PQAR document [D4].
Figure 4-1: Results reproduced from [D1]; Top: Validation results for Cloud_cci surface incoming shortwave flux using BSRN as a reference. Bottom: Bias for each ground station.
Validation of BOA fluxes against BSRN stations present standard deviations of 24 W/m² and bias of 8.2 W/m² for Surface Incoming Shortwave radiation (SIS) and standard deviation of 14 W/m² and bias of 11.9 W/m² for Surface Downwelling Longwave radiation (SDL). The intercomparison of Cloud_cci radiation products with CERES present a bias of 1.53 W/m², standard deviation of 3.18 W/m² and stability of 0.97 W/m2/decade for SIS. Bias of 10.17 W/m², standard deviation of 1.2 W/m² and stability of 2.8 W/m2/decade for SDL. Intercomparison (using the monthly mean data from January 2017 to December 2021) of ICDR products with CERES showed biases (we report here the maximum between the values find for SLSTR-A and SLSTR-B) consistent with TCDR and are: 1.5 W/m² for SIS, 2.0 W/m² for SRS, 3.9 W/m² for SOL and 13 W/m² for SDL.
References
Loeb, N.G., Doelling, D.R., Wang, H., Su, W., Nguyen, C., Corbett, J.G., Liang, L., Mitrescu, C., Rose, F.G., and Kato, S.: Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) Top-of-Atmosphere (TOA) Edition 4.0 Data Product, J.Climate, 31(2), 895–918, doi:10.1175/JCLI-D-17-0208.1, 2018.
Poulsen, C. A., McGarragh, G. R., Thomas, G. E., Stengel, M., Christensen, M. W., Povey, A. C., Proud, S. R., Carboni, E., Hollmann, R., and Grainger, R. G.: Cloud_cci ATSR-2 and AATSR data set version 3: a 17-year climatology of global cloud and radiation properties, Earth Syst. Sci. Data, 12, 2121–2135, 2020, https://doi.org/10.5194/essd-12-2121-2020.