Contributors: D. Ghent (UoL), K. Veal (UoL), G.Kirches (BC)
Issued by: University of Leicester/ Darren Ghent
Date: 07/10/2025
Ref: C3S2_313e_BC_WP3-DR-LST-LST_CCI-v3.00-1995-2024_202506_PUGS
Official reference number service contract: 2024/C3S2_313e_BC/SC1
History of modifications
List of datasets covered by this document
Acronyms
General definitions
Brightness temperature (BT): Brightness temperature is the temperature of a black body that would have the same radiance as the radiance actually observed with a radiometer. It is a directional temperature obtained by equating the measured radiance with the integral over wavelength of the Planck’s Black Body function times the sensor response. The relation between the directional radiance RB,i(θφ) and the target brightness temperature TB,i(θφ) is given by
\[ R_{B,i}(\theta,\phi)=R(T_{B,i} (\theta,\phi))=\int_{\lambda_1}^{\lambda_2} \frac {f_i(\lambda)C_1} {\pi {\lambda^5} \left[ \exp \left( \frac {C_2} {\lambda T_{B,i}(\theta, \phi)} \right) -1 \right]} \;d{\lambda} \]
where λ1 and λ2 are the lower and upper wavelength limits of the sensitivity of channel i, C1 = 3.7404 x 108 W μ4 m-2, C2 = 14387 μ K, and fi(λ) is the spectral response of channel i between λ1 and λ2 (Norman and Becker, 1995).
Brokered dataset: A dataset whose development and processing were not funded by C3S but is now distributed by C3S.
Calibration: Calibration is the process of quantitatively defining the system response to known, controlled system inputs (Guillevic et al., 2018).
Equator Crossing Time (EXT): local solar time when satellite passes over the equator
Error: Result of a measurement minus a true value of the measurand. Note that in practice a true value cannot be determined and therefore a conventional true value is used instead (JCGM, 2008).
Land surface temperature (LST): Land Surface Temperature (LST) is the aggregated radiative skin temperature derived from thermal radiation of all objects comprising the surface, measured in situ and estimated from satellite. It is a basic determinant of the terrestrial thermal behaviour, as it controls the effective radiating temperature of the Earth’s surface (Norman and Becker, 1995; Guillevic et al., 2018).
Level-1 (L1b): radiometrically calibrated and geometrically corrected radiances or brightness temperatures presented on the orbit swath at native resolution and geolocated to latitude and longitude of centres (and/or corners) of pixels or to tie-points. Should the geolocation information be given at tie-points only, then the user is required to perform an interpolation in order to geolocate pixels in between the tie-points.
Level-2 (L2): geophysical variables, e.g. LST, derived from L1 data at same resolution as L1 data i.e. not spatially or temporally manipulated. Geolocation information may be included with the L2 product, but not necessarily at full resolution, or in other cases the user must source the geolocation information from a L1 product.
Level 3 (L3): geophysical product that has been temporally or spatially manipulated and in a gridded map projection format e.g. daily LST on a 0.01° longitude by 0.01° latitude global grid.
Level-3 uncollated (L3U): L2 data regridded to a spatial grid without combining data from different orbits.
Level-3 collated (L3C): L2 data from a single instrument regridded to a space-time grid, data from several orbits may be combined.
Level-3 super-collated (L3S): L3 data from multiple instruments combined in a space-time grid.
Skin temperature: The temperature of the surface layer of a medium of depth equal to the penetration depth of the electromagnetic radiation at the given wavelengths (Norman and Becker, 1995). In the case of land surface temperature retrieved using infra-red radiation at wavelengths around 10 microns the skin has a depth of a few microns to less than a millimeter.
Split-Window (SW): Refers to the use of adjacent infra-red bands to correct for atmospheric effects based on differential absorption (Wan and Dozier, 1996).
Total column water vapour (TCWV): Amount of water (depth of vertical column of unit-crossectional area) which would be obtained if all the water vapour in a specified column of the atmosphere were condensed to liquid. The equivalent term precipitable water (PW) is also frequently used.
Uncertainty: A parameter associated with the result of a measurement, that characterizes the dispersion of the values that could reasonably be attributed to the measurand, that is the value of the particular quantity to be measured (JCGM, 2008).
Executive summary
The Land Surface Temperature (LST) Climate Data Record (CDR) and Interim Climate Data Record (ICDR) uses or will use data from a series of four satellite instruments to produce a climate data record with a temporal span approaching 33 years (from 1995 to 2027). Of the four instruments, three have a common European Space Agency (ESA) heritage: the Along-Track Scanning Radiometer 2 (ATSR-2) on Earth European Remote-sensing Satellite - 2 (ERS-2), the Advanced Along-Track Scanning Radiometer (AATSR) on Environmental Satellite (Envisat) and the Sea and Land Surface Temperature Radiometer (SLSTR) on Sentinel-3B (successors Sentinel-3C/D). The fourth is the Moderate Imaging Spectroradiometer on Earth Observation System (EOS) - Terra (MODIS Terra). The dataset provides monthly day and night LSTs in separate netCDF files on 0.01° x 0.01° longitude-latitude grids. Per pixel uncertainty estimates are given as an estimate of the total uncertainty and also as a breakdown of the uncertainty into components by correlation length, allowing users to propagate uncertainties through spatial and temporal averaging. In order to produce a consistent time series, calibration differences between the four instruments have been addressed and adjustments for the different equator crossing times (EXTs) of the instruments have been applied. Both the calibration and EXT time difference adjustments are made to the brightness temperatures before retrieving the land surface temperature. The processing chain applies a generalised split-window retrieval algorithm to Level 1B brightness temperatures and outputs day and night LSTs on separate 0.01° x 0.01° longitude latitude grids. Users should be aware that this product contains clear-sky LSTs and that monthly fields will contain spatial gaps in the case of persistent cloud. There are also gaps in the time series (January 1996 - June 1996 and February 2001 - June 2001) due to instrument and platform issues. Although efforts have been made to remove inter-instrument differences, small signals may remain in the time series at instrument change points and should not be attributed to climate phenomena.
1. Product Description
Land Surface Temperature (LST) describes processes such as the exchange of energy and water between the land surface and the atmosphere, and influences the rate and timing of plant growth. LST, which has been defined as a Global Climate Observing System (GCOS) Essential Climate Variable (ECV), is increasingly being used in diverse applications such as drought monitoring and crop management, hydrological monitoring and water management, investigation of evapotranspiration, land-atmosphere feedbacks, and landcover change, climate modelling and data assimilation, numerical weather prediction, forecasting and reanalysis, permafrost monitoring, and the study of urban temperatures. The dataset described here (ESA CCI LST Infra-red Climate Data Record version 3.00, CDR LST_CCI LST v3.00) provides a near 30-year record (1995 to 2024) of consistently processed LST data in which inter-instrument differences have been addressed. The Interim Climate Data record will extend the dataset to 2027.
1.1. Land Surface Temperature product description
1.1.1. Overview
The ESA CCI IRCDR V3.00 (brokered as CDR LST_CCI LST v3.00) contains land surface temperatures (LSTs) and their uncertainty estimates from multiple Infra-Red (IR) instruments on Low Earth Orbiting (LEO) sun-synchronous (polar orbiting) satellites. LSTs are provided on a global equal angle grid at a resolution of 0.01° longitude and 0.01° latitude. Surface temperatures are given over all cloud-free land and sea-ice (see section 1.4.5 for more information on validation of LSTs over snow and ice surfaces). Daytime and night-time temperatures are provided in separate files corresponding to 10:00 and 22:00 local solar time. The CDR will be updated as an Interim Climate Data Record ICDR LST_CCI LST v3.00 2025-2027.
1.1.2. Instruments
The CDR dataset is comprised of LSTs from four instruments. Three of the instruments have a common European Space Agency (ESA) heritage: the Along-Track Scanning Radiometer 2 (ATSR-2) on Earth European Remote-sensing Satellite - 2 (ERS-2), the Advanced Along-Track Scanning Radiometer (AATSR) on Environmental Satellite (Envisat), and the Sea and Land Surface Temperature Radiometer (SLSTR) on Sentinel-3B. The fourth is the Moderate Imaging Spectroradiometer on Earth Observation System (EOS) - Terra (MODIS Terra) which is used to fill the gap between the end of the AATSR mission and the start of operational delivery from Sentinel-3B. Thus, the instruments contributing to the time series are: ATSR-2 from June 1995 to May 2002; AATSR from June 2002 to March 2012; MODIS Terra from April 2012 to November 2018; and SLSTR from December 2018 to December 2024. The Interim Climate Data Record - ICDR LST_CCI LST v3.00 2025-2027 will be based on SLSTR data from January 2025 to December 2027.
Figure 1: Time coverage for each of the sensors used in the ESA CCI IRCDR V3.00 (1995-2024) brokered as CDR LST_CCI LST v3.00. Note, that there are gaps in the ATSR-2 time series due to instrument (January - June 1996) and platform (February - June 2001) anomalies.
1.1.3. Algorithm
For a full description of the input datasets and algorithm see the Algorithm Theoretical Basis Document (ATBD). The basic characteristics of the algorithm are provided in Table 1 and an outline of the algorithm is given subsequently.
Table 1: Algorithm History
| Algorithm Name | Version | Type | History |
|---|---|---|---|
| Multisensor IRCDR | 3.00 | Generalised Split-Window | Product brokered from ESA CCI |
The input data are the Level 1B brightness temperatures from the four instruments mentioned above. The different instruments used for this dataset have different swath widths, for example AATSR has a swath width of 500 km and MODIS has a swath width of 2,330 km. For consistency through the time series, data coverage of all instruments is restricted to the width of the narrowest swath (500 km).
Inter-instrument biases are accounted for by cross-calibration with the Infrared Atmospheric Sounding Interferometer (IASI) instruments on Meteorological Operational (METOP) satellites. Furthermore, an adjustment is made to the LSTs to account for the half-hour difference between satellite equator crossing times: the local equator crossing times (EXTs) for Envisat and Sentinel-3B are 10:00 and 22:00 whereas the EXTs for ERS-2 and EOS-Terra are 10:30 and 22:30. The LSTs for ERS-2 ATSR-2 and EOS-Terra MODIS are adjusted to 30 minutes earlier than the observation time. A radiative transfer model is used to simulate the brightness temperatures at the observation time and 30 minutes earlier. The difference is applied to the actual observed brightness temperature and the adjusted brightness temperatures are then propagated through the retrieval algorithm to derive the LST at the earlier time.
The Generalised Split-Window (GSW) algorithm is used to retrieve the LST for each sensor. The generalised split-window algorithm is a view-angle dependent split-window algorithm proposed for LST retrieval by Wan and Dozier (1996). It is based around channels in the 11 and 12 µm regions. The algorithm coefficients are derived from linear regression using data simulated with the RTTOV (Radiative Transfer for TOVS, Saunders et al., 2018) radiative transfer model. The coefficients are dependent on satellite viewing angle and water vapour. Emissivities are taken from a climatology of the Combined ASTER and MODIS Emissivity for Land (CAMEL) V2 database (Borbas et al., 2018).
A consistent cloud detection method is applied to all sensors contributing to the dataset. This algorithm (named UOL_3) is a semi-Bayesian cloud masking approach using the probability of clear-sky conditions, which has been developed at University of Leicester (Ghent et al., 2017). A pixel-level cloud mask is derived using a combination of simulated brightness temperatures and observational climatology. The approach is equally valid for both day and night-time retrievals as this method is independent of visible wavelength information. LSTs are not produced where clouds are present since under these circumstances the IR radiometer observes the cloud top which is usually much colder than the land surface.
LST data are provided as monthly averages which are calculated as arithmetic means of available daily values. The uncertainties are propagated using the methods described in Bulgin et al. (2023).
1.1.4. Coverage
The dataset coverage is global over the land surface and sea-ice. During the period covered by ATSR-2, small regions were not covered due to downlinking constraints (most noticeably a track extending southwards across central Asia through India) – further details can be found on the ATSR project webpages.
1
Dataset coverage starts on 1st June 1995 and ends on 31st December 2024. There are two gaps of several months in the dataset. Firstly, no data were acquired from ATSR-2 between the 23rd December 1995 and 30th June 1996 due to a scan mirror anomaly. Secondly, the ERS-2 gyroscope failed in January 2001 so that data quality was poor due to geolocation errors between 17th Jan 2001 and 5th July 2001 and these data are not used in this dataset. Also, there is a twelve day gap in the dataset due to Envisat mission extension orbital manoeuvres from 21st October 2010 to 1st November 2010. There are minor interruptions (1-10 days) during satellite/instrument maintenance periods or instrument anomalies.
1.1.5. Data delivery
The dataset was produced by the University of Leicester (UoL) in the UoL processing chain as part of the ESA Land Surface Temperature Climate Change Initiative, which strives to improve satellite datasets to Global Climate Observing System (GCOS) standards.
The CDR will be updated as an Intermediate Climate Data Record (ICDRs) at six monthly intervals in June and December of each year with a latency of 6 months.
1.2. Overview of Product Target Requirements
The dataset is assessed against the GCOS-245 (WMO, 2025) requirements for LST (see section 5 of the Product Quality Assessment Report for details), results are summarised in Table 2.
Table 2: Compliance with GCOS Requirements: G is Goal, B is Breakthrough, and T is target value.
Requirement | GCOS-245 Requirement | Reported value | ||
G | B | T | ||
Horizontal Resolution (km) | < 1 | < 1 | 1 | 0.01° (~ 1km, within Goal) |
Temporal Resolution (h) | < 1 | 1 | 6 | Day / Night (> 6 hours, does not meet Threshold) |
Timeliness (d) | 2 | 30 | 1 month (~ 30 days, within Threshold) | |
Required Measurement Uncertainty (K) | < 1 | < 1 | < 1 | < 1 K (within Goal) |
Stability (K / decade) | 0.1 | 0.2 | 0.3 | 0.11 (meets BreakThrough) |
This LST dataset has a horizontal resolution of 0.01° longitude and latitude which approximately meets the GCOS requirement of < 1 km depending on location over the globe. LSTs are provided for daytime and nighttime, thus at 12 hourly intervals so the GCOS threshold for temporal resolution of 6 hourly is not met. Only geostationary satellites can provide data at these temporal resolutions but these are regional datasets only, whereas polar orbiting satellites cover the whole globe but are restricted to day/night temporal resolution. However, the notes to the GCOS-245 temporal resolution requirement state that the day/night temporal resolution from polar orbiting satellites satisfies 70% of climate users (GCOS, 2025).
The GCOS requirement on timeliness was the outcome of a survey of 80 users, which identified a “threshold” need of 30 days for long-term data records, and a “breakthrough” of 2 days for long-term data records. The C3S LST products (CDR LST_CCI LST v3.00 1995-2024) were produced with a timeliness of 30 days, thus meeting the threshold requirement. The ICDR will not meet the criteria as updates will be made every 6 months.
The required measurement uncertainty is the total uncertainty per pixel combining the four groups of uncertainty components: random, locally correlated atmospheric, locally correlated surface, and large scale systematic (as described in the Algorithm Theoretical Basis Document of the C3S LST Products, E.U. Copernicus Climate Change Service, 2025). The uncertainties are validated and are within 1 K thus meeting the GCOS goal.
For the C3S LST products (CDR LST_CCI LST v3.00 1995-2024) a stability value of 0.11 K/decade was estimated, thus almost meeting the goal requirement.
1.3. Key variables
1.3.1. Land surface temperature
The land surface temperature data are split into day and night products using the solar zenith angle (daytime when solar zenith angle < 90° and nighttime when solar zenith angle > 90°). For this reason, daytime temperatures during polar night and nighttime temperatures during polar day are not provided by this product. For example, in Figure 2 white areas correspond to pixels with no available LST. There are no daytime temperatures over Antarctica in June (Figure 2a) however there are nighttime LSTs (Figure 2b) for the same month. The reverse is true for the Arctic in June (Figure 2a and 2b) and for Antarctica in December (Figure 2c and 2d). Other areas with no data are regions of persistent cloud cover as well as lakes and ice-free ocean where LSTs are not processed.
(a)
| (b)
|
(c)
| (d)
|
Figure 2: Monthly land surface temperature in June 2024 for (a) daytime and (b) nighttime and in December 2024 for (c) daytime and (d) nighttime.
1.3.2. Total uncertainty and uncertainty components
Uncertainties are given in two forms:
an estimate of the total uncertainty for the pixel
a breakdown of the uncertainty into components by correlation length (Ghent, 2019).
The total uncertainty is the sum in quadrature of the individual components. Figure 3a shows the total uncertainty on the monthly daytime LST for June 2024.
The components of the uncertainty budget are:
uncertainty due to uncorrelated or "random" errors
uncertainty due to errors correlated on surface scales
uncertainty due to errors correlated on atmospheric scales
uncertainty due to locally correlated errors on LST corrections
uncertainty due to correlated or "systematic" errors
The uncertainty due to uncorrelated errors (Figure 3b) has contributions from instrument noise and also from uncorrelated errors in the emissivity which is allocated to the pixel. In addition, this component has a contribution from temporal under-sampling since on some days there will not be a valid LST, for example, if there is cloud cover. Errors correlated on atmospheric scales include errors in the assumed total column water vapour which is taken from ERA 5 (Hersbach et al., 2023). The uncertainty from errors correlated on atmospheric scales for June 2024 is plotted in Figure 3c. As expected, the uncertainty is greatest over regions of high atmospheric water vapour. Figure 3d shows the uncertainty due to errors correlated on surface scales. These are the correlated errors in the surface emissivity used in the LST retrieval. The uncertainty due to simulation model error (not plotted) is a uncorrelated error, with a value of approximately 0.029 K, small in respect to the other uncertainty components.
(a)
| (b)
|
(c)
| (d)
|
Figure 3: Uncertainty on the monthly daytime LST for June 2024: (a) total uncertainty, (b) uncertainty due to uncorrelated errors, (c) uncertainty due to errors correlated on atmospheric scales, (d) uncertainty due to errors correlated on surface scales.
LSTs from ATSR-2 and Terra MODIS have an adjustment applied to account for the overpass times of the sensors, which is 30 minutes later than those of AATSR and Sentinel 3B SLSTR (see Section 1.1 and the ATBD). This adjustment is calculated using a radiative transfer model to simulate the difference between the brightness temperature at the observation time and the brightness temperature 30 minutes earlier. This difference is then applied to the brightness temperatures before retrieving the LST. The uncertainty in the brightness temperature adjustment is propagated through the retrieval to calculate the additional uncertainty in the LST that is due to the adjustment. The uncertainty tends to be higher over regions with higher diurnal range and higher probability of cloud contamination (Figure 4). Note, that for AATSR and SLSTR, which are taken as the reference time and therefore not adjusted, this variable is set to zero.
Figure 4: Uncertainty on monthly daytime LST due to errors in time correction for June 2016.
1.4. Data usage information
1.4.1. Data format and file naming
The data are supplied in netCDF4 files and are compliant with CF-conventions v1.8 and CCI Data Standards v2.2 (ESA Climate Office, 2018).
The file naming follows the ESA Climate Office CCI data standards (version 2.0) :
ESACCI-LST-<Processing Level>-LST-<Product String>[-<Additional Segregator>]-<Indicative DateTime>-fv<FileVersion>.nc
<Processing Level>
The data processing level code which is L3S for this product.
<Product String>
The product string is IRCDR_ for this dataset.
[-<Additional Segregator>]
The additional segregator for this product indicates the spatial and temporal resolution and whether the product contains daytime or nighttime temperatures. For example, 0.01deg_1MONTHLY_NIGHT indicates a 1 month composite of nighttime data at 0.01° resolution and 0.01deg_1MONTHLY_DAY indicates a 1 month composite of daytime data at 0.01° resolution.
<Indicative DateTime>
The identifying datetime for this data set. The format is YYYYMMDDhhmmss, where YYYY is the four digit year, MM is the two digit month and DD is the two digit day of the month. The time of day, hhmmss, is not applicable for the monthly files and is always given as 000000. The date identifies the first day of the composite period.
<FileVersion>
File version number in the form n{1,}[.n{1,}] (That is 1 or more digits followed by optional . and another 1 or more digits.). This dataset has version number 3.00.
As an example, the file for July 2006 for this dataset has the name:
ESACCI-LST-L3S-LST-IRCDR_-0.01deg_1MONTHLY_DAY-20060701000000-fv3.00.nc
1.4.2. Quality flags and data masks
There are no quality flags in the files as the quality flags were applied at earlier processing levels (see the ATBD).
1.4.3. File contents
Table 3 shows the variables in a single product file. Table 4 shows the global attributes of a single product file. A list of variables, variable attributes, and global attributes obtained using the ncdump utility is provided in the Annexe.
Table 3: Variables contained in the LST product files
| variable name | variable long name | Comment |
|---|---|---|
| time | reference time of file | |
| dtime | time difference from reference time | Offset from reference time in seconds. For time adjusted data this may be negative |
| lat | latitude_coordinates | |
| lon | longitude_coordinates | |
| satze | satellite zenith angle | |
| sataz | satellite azimuth angle | |
| solze | solar zenith angle | |
| solaz | solar azimuth angle | |
| lst | land surface temperature | |
| lst_uncertainty | land surface temperature total uncertainty | |
| lst_unc_ran | uncertainty from uncorrelated errors | |
| lst_unc_loc_atm | uncertainty from locally correlated errors on atmospheric scales | |
| lst_unc_loc_sfc | uncertainty from locally correlated errors on surface scales | |
| lst_unc_sys | uncertainty from large-scale systematic errors | |
| lcc | land cover class | |
| n | number of clear-sky pixels | Number of days contributing to the monthly average |
| channel | channel wavelength in microns | |
| lst_unc_loc_cor | uncertainty from locally correlated errors on LST corrections | Uncertainty in LST due to time adjustment. This variable has a value of zero where no adjustment is applied (AATSR, Sentinel 3b SLSTR) |
Table 4: File global attributes
| attribute name | attribute value |
|---|---|
| title | "ESA LST CCI land surface temperature data at product level L3C from Multi-sensor." ; |
| institution | "University of Leicester" ; |
| history | "Created using software developed at University of Leicester" ; |
| references | "https://climate.esa.int/en/projects/land-surface-temperature" ; |
| Conventions | "CF-1.8" ; |
| product_version | "3.00" ; |
| keywords | "Earth Science, Land Surface, Land Temperature, Land Surface Temperature" ; |
| id | "ESACCI-LST-L3S-LST-IRCDR_-0.01deg_1MONTHLY_DAY-20030101000000-fv3.00.nc" ; |
| naming_authority | "le.ac.uk" ; |
| keywords_vocabulary | "NASA Global change Master Directory (GCMD) Science Keywords" ; |
| cdm_data_type | "grid" ; |
| comment | "These data were produced as part of the ESA LST CCI project." ; |
| date_created | "20250616T095533Z" ; |
| creator_name | "University of Leicester Surface Temperature Group" ; |
| creator_url | "https://climate.esa.int/en/projects/land-surface-temperature" ; |
| creator_email | "djg20@le.ac.uk" ; |
| project | "Climate Change Initiative - European Space Agency" ; |
| geospatial_lat_min | -89.995f ; |
| geospatial_lat_max | 89.99499f ; |
| geospatial_lon_min | -179.995f ; |
| geospatial_lon_max | 179.995f ; |
| geospatial_vertical_min | 0.f ; |
| geospatial_vertical_max | 0.f ; |
| time_coverage_start | "20030101T000000" ; |
| time_coverage_end | "20030131T235959" ; |
| time_coverage_duration | "P1M" ; |
| time_coverage_resolution | "P1M" ; |
| standard_name_vocabulary | "CF Standard Name Table v71" ; |
| license | "ESA CCI Data Policy: free and open access" ; |
| platform | "Multi-platform" ; |
| sensor | "IRCDR_" ; |
| geospatial_lat_units | "degrees_north" ; |
| geospatial_lon_units | "degrees_east" ; |
| geospatial_lon_resolution | 0.01f ; |
| geospatial_lat_resolution | 0.01f ; |
| key_variables | "land_surface_temperature" ; |
| doi | "10.5285/b8be3f3993e54d309fddf61ea5d3674f" ; |
| code_version | "3eb2ad3b7d75dc845aa6704d127da63bb815c5e9" ; |
| format_version | "CCI Data Standards v2.2" ; |
| spatial_resolution | "0.01 degree" ; |
| source | "ESA LST CCI ATSR_3 L3S V3.00" ; |
| summary | "This file contains level L3S global land surface temperatures from AATSR. L3S data are derived from more than one sensor combined on a space and/or time grid." ; |
1.4.4. Examples of known climate applications and best practices
This dataset can be used for climate monitoring and assessment of climate models. Land surface temperatures are radiative skin temperatures (see General Definitions above) which are different from surface air temperatures. Surface air temperatures are usually measured at a distance above the surface, usually with a mercury thermometer. The land surface temperatures are retrieved from infra-red data in clear-sky only conditions, with the result that data gaps are present in the case of persistent cloud cover.
Advice on handling the uncertainty information when averaging in space can be found in the appendix to Bulgin et al. (2023). In brief, the error correlations must be considered when propagating the uncertainties. Uncorrelated errors reduce on averaging whereas correlated errors do not. In the case of the ESACCI LST IRCDR, the locally correlated components (variable names lst_unc_loc_atm, lst_unc_loc_sfc, and lst_unc_loc_cor) are assumed to be correlated on scales up to 5km (~0.05°) and assumed uncorrelated at larger scales. For example, if we wish to regrid the 0.01° LST to a 0.25° grid using an arithmetic mean, we can propagate each locally correlated component first to 0.05°, treating it as correlated (correlation coefficient equal to one), and then propagate it from 0.05° to 0.25°, treating it as uncorrelated. The ESACCI Land Surface Temperature Project has produced a tool that can be used to subset and regrid ESACCI LST products, propagating the uncertainty components as recommended above. The tool and accompanying manual are available to download from the “Tool” tab on the ESACCI Land Surface Temperature Project website. The ATBD for the tool (Consortium CCI LST, 2023) is available on the “Key Documents” tab at the same site.
1.4.5. Known Issues and Limitations
Although efforts have been made to remove inter-instrument differences, small signals may remain in the time series at instrument change points and should not be attributed to climate phenomena.
| Issue name | Description of Issue | What to do |
|---|---|---|
| Lack of validation of LST over ice/snow surfaces | There is a paucity of reference LST in situ data over ice and snow surfaces so no validation of the data over these surfaces is provided in the PQAR. However, the same algorithm is used over all surfaces and no impact on performance is expected over ice. One year of validation in ESA LST_cci of the brokered dataset indicates it performs better over ice than equivalent LST datasets - see Section 2.2 (Impact of CCI LST IST products from MODIS and SLSTR on the Arctic SST/IST Multi-Year (MY) Product of the Copernicus Marine Service (Ioanna Karagali, Adrien Combelles and Pia Englyst, DMI)) in LST_cci-D5.1-CAR_-_i3r1_-_Phase2_CAR. | If users wish to mask out the sea-ice or land ice/snow data they should use the land cover classification data (variable ‘lcc’) in the product. Sea-ice pixels are flagged with a value of 230 and land permanent snow/ice are flagged with 220. |
| Stability of early period of record | There are fluctuations in the mean bias in the early part of the record (years 1995-1999). Results of a stability analysis against CRUTEM5 air temperature anomalies show the record to be GCOS compliant over the whole period with a linear trend of 0.11 K /decade. However, for the years after 1999 the stability is improved and the linear trend is 0.03 K/decade. A description of the stability analysis, with results, can be found in the PQAR Section 5. | For climate trend analysis, it is recommended that users exclude data before the start of 2000. |
2. Data access information
The data can be accessed through the CDS using this link: https://cds.climate.copernicus.eu/ and searching for Land Surface Temperature. The product Data Object Identifier (DOI) is https://dx.doi.org/10.5285/b8be3f3993e54d309fddf61ea5d3674f. The licence for use is the ESA CCI Data Policy: free and open access.
3. User Support
Users can raise a query regarding this dataset on the ECMWF Support Portal.
References
Borbas, E. E., G. Hulley, M. Feltz, R. Knuteson and S. Hook (2018). The Combined ASTER MODIS Emissivity over Land (CAMEL) Part 1: Methodology and High Spectral Resolution Application. Remote Sensing 10(4): 643. https://doi.org/10.3390/rs10040643
Bulgin, C., S. Ermida and C. Jimenez (2023). End-To-End ECV Uncertainty Budget, V3.0. p.82-83. Under Key Documents at https://climate.esa.int/en/projects/land-surface-temperature/
Consortium CCI LST (2023). Re-gridding and Sub-setting ATBD. Under Key Documents at https://climate.esa.int/en/projects/land-surface-temperature/
ESA Climate Office (2018). CCI data standards 2.3. CCI-PRGM-EOPS-TN-13-0009 Available from https://climate.esa.int/media/documents/CCI_DataStandards_v2-3.pdf
E.U. Copernicus Climate Change Service (C3S) (2025). Algorithm Theoretical Basis Document (ATBD) v1.0 - CDR and ICDR Land Surface Temperature v3.00 C3S2_313e_BC_WP3-DR-LST-LST_CCI-v3.00-1995-2024_202506_ATBD_v1.0
Ghent, D., K. Veal, T. Trent, E. Dodd, H. Sembhi and J. Remedios (2019). A New Approach to Defining Uncertainties for MODIS Land Surface Temperature. Remote Sensing 11(9): 1021. https://doi.org/10.3390/rs11091021
Ghent, D. J., G. K. Corlett, F. M. Gottsche and J. J. Remedios (2017). Global Land Surface Temperature From the Along-Track Scanning Radiometers. Journal of Geophysical Research-Atmospheres 122(22): 12167-12193.
Guillevic, P., F. Göttsche, J. Nickeson, G. Hulley, D. Ghent, Y. Yu, I. Trigo, S. Hook, J. Sobrino, J. Remedios, M. Román, F. Camacho (2018). Land Surface Temperature Product Validation Best Practice Protocol. Version 1.1. https://lpvs.gsfc.nasa.gov/PDF/CEOS_LST_PROTOCOL_Feb2018_v1.1.0_light.pdf
Hersbach, H., B. Bell, P, Berrisford, G. Biavati, A. Horányi, J. Muñoz Sabater, J. Nicolas, C. Peubey, R. Radu, I. Rozum, D. Schepers, A. Simmons, C. Soci, D. Dee, J-N. Thépaut (2023). ERA5 hourly data on pressure levels from 1940 to present. Copernicus Climate
Change Service (C3S) Climate Data Store (CDS), DOI: 10.24381/cds.bd0915c6
Joint Committee for Guides in Metrology (JCGM, 2008). Evaluation of Measurement Data - Guide to the Expression of Uncertainty in Measurement. https://doi.org/10.59161/JCGM100-2008E
Norman, J. M. and F. Becker (1995). Terminology in thermal infrared remote sensing of natural surfaces. Agricultural and Forest Meteorology 77(3): 153-166. https://doi.org/10.1016/0168-1923(95)02259-Z
Saunders, R., J. Hocking, E. Turner, P. Rayer, D. Rundle, P. Brunel, J. Vidot, P. Roquet, M. Matricardi, A. Geer, N. Bormann and C. Lupu (2018). An update on the RTTOV fast radiative transfer model (currently at version 12). Geosci. Model Dev. 11(7): 2717-2737. https://doi.org/10.5194/gmd-11-2717-2018
Wan, Z. and J. Dozier (1996). A generalized split-window algorithm for retrieving land-surface temperature from space. IEEE Transactions on Geoscience and Remote Sensing 34(4): 892-905. doi: 10.1109/36.508406
World Meteorological Organization (2025). The 2022 GCOS ECV Requirements. GCOS-245. Geneva: World Meteorological Organization. https://library.wmo.int/idurl/4/58111
Annexe A
Below is the output from ncdump of one of the LST product files (ESACCI-LST-L3S-LST-IRCDR_-0.01deg_1MONTHLY_DAY-20180101000000-fv3.00) showing the file dimensions, variables and associated attributes, and the file global attributes.
