ERA-CLIM2 project activities have been organised in six scientific work packages:

WP1 - Global 20th century reanalysis

Two types of reanalyses have been produced using the ECMWF coupled assimilation system: an extended climate reanalysis of the 20th century at moderate resolution (CERA-20C, covering the 110-year period 1901-2010), and a higher-resolution reanalysis of few years of the satellite era suitable for near-real time climate monitoring (CERA-SAT, covering the period 2008-2016). Consistent estimates of the carbon cycle (land and ocean) have been produced for both sets of products. Ensemble techniques have been used in reanalysis production to provide uncertainty assessments. All input observations and associated quality feedback information have been provided to users via the ERA-CLIM Observation Feedback.


Figure 1. MSLP reanalyses' comparison (see text).

As an example of the WP1 deliverables, figure 1 shows time series of mean sea level pressure (MSLP) for the latitudes 90°S–60°S averaged over the period September-November each year in one model integration (ERA-20CM) and various reanalyses (ERA-20C, NOAA 20CR, CERA-20C). The trend in the ERA-CLIIM2 CERA-20C looks more realistic without the spurious drift present in ERA-20C and 20CR. This has been linked to the better specification of background and observation errors in CERA-20C.

WP2 - Future coupling methods

Research and development in coupled data assimilation for climate reanalysis have been conducted, targeted for implementation in the ECMWF coupled assimilation system used in the reanalyses production (the CERA system), or in experimental assimilation systems that could be used in future productions (e.g. by the Copernicus Climate Change Service). This includes work on improving the use of surface observations in coupled data assimilation, on improving various aspects of the ocean analysis component, on improving the carbon component of a coupled earth system reanalysis, and longer term research toward fully coupled data assimilation. ERA-CLIM2 work contributed to advancing and testing coupling techniques, e.g. between the 3-dimensional ocean and the atmosphere (Fig. 2).

Figure 2. Lead-lag correlations between SST and total precipitation in a region in the western Tropical Pacific. The left plot shows results from ERA-20C uncoupled reanalysis while the right plot shows results from the ERA-CLIM2 CERA-20C coupled reanalysis. In both plots, the dashed line is the results from the 1900s while the solid colored lines are from the 2000s, and the black solid lines are observational estimates.

WP3 - Earth system observations

Major work on recovery and preparation of input observations for climate reanalysis has taken place in this work package. Activities in satellite data rescue, reprocessing and recalibration, will contribute to improvement of the satellite record and its extension back to the 1960’s. In-situ data rescue activities started in ERA-CLIM have continued and have been expanded. Work on improving the input data sets needed to constrain sea-surface temperature, sea-ice, and snow cover will be included. Figure 3 above shows a geographical map of the regions where global in-situ data rescue activities were conducted in ERA-CLIM2.

https://sites.google.com/site/eraclim2/work-packages/data_rescue_activities-large.png?attredirects=0
Figure 3. ERA-CLIM2 global in-situ data rescue activities.

Considering, for example, surface meteorological data, observations of pressure and wind and, sometimes, temperature are assimilated into reanalyses. ERA-CLIM2 contributed to the collection of further historical surface meteorological data by cataloguing, prioritising, imaging, digitising, applying quality control tools and/or formatting sub-daily data inform from various sources (see Table 1, from Brönnimann et al, 2018, BAMS). Thanks to the two EU-FP7 projects ERA-CLIM and ERA-CLIM2, about 5.5 million station-days of surface observations have been catalogued and digitized (Table 1; of them, 2.2M were catalouged and digitized within the ERA-CLIM2 project). A similar work has been done for upper-air data (1.3M station-days catalogued and digitized by the ERA-CLIM and ERA-CLIM2 projects, of which 0.7M by ERA-CLIM2; see Brönnimann et al, 2018, BAMS, for more details).
 

 Source Cataloged Digitized Quality Controlled
Backward extension (<1965) of meteorological data from 246 Russian stations  2,738,595 2,738,595 2,738,595
41 Chilean stations 1950-1999
 383,151 357,456 36,682
76 Portuguese stations in Portugal and ex-colonies in Africa and Asia 1,020,727 1,009,131 605,478
South China Sea logbooks for 100 stations
 830,286 830,286 830,286
Snow data for 20 stations in Russia
 622,325 622,325 622,325
Table 1. Surface observations (in station days) digitized within ERA-CLIM and ERA-CLIM2 (see Broennimann et al, 2018, BAMS).

WP4 - Quantifying and reducing uncertainties

Goals for this work package included making optimal use of observations in reanalysis, and providing end users with meaningful information about uncertainties in reanalysis products. This involved a range of activities, including quality control and error estimation for input observations, work on bias correction and homogenisation of data records, and various quality assessment of reanalysis products based on independent observations and comparisons with other reanalyses and high-level observational products. An example is given in Fig. 4, which shows the time evolution of the uncertainty in the reanalysis value for the annual average mean-sea-level-pressure (MSLP): the red bars show how the uncertainty in the average MSLP value reduces throughout the 20th century, since an increasing number of more accurate observations becomes available.

Figure 4. Time evolution of the annual-average mean-sea-level-pressure over the Southern Hemisphere between 60-90 degrees South: the solid black line shows the CERA-20C ensemble-average and the red bars the ensemble standard deviation. As time progresses, more accurate observations become available, and this reduces the uncertainty, represented by the size of the red bars.

WP5 - Data service developments

This work package addressed development of efficient and sustainable data and web services that are essential to the provision of full and open access to the data and information products developed in the project. Thanks to this work, for example, the two coupled reanalyses CERA-20C and CERA-SAT can be accessed usign the ECMWF MARS system (Fig. 5).

Figure 5. MARS interface to access the 110 years of the coupled CERA-20C reanalysis ensemble produced by ERA-CLIM2.

WP6-9 - Dissemination and outreach, project management and coordination with EU-FP7 sister projects

WP6 work package was concerned with ensuring effective dissemination of project products (data, graphics, and reports) and with facilitating the uptake of these products by future end users of climate data. It also included work aimed to coordinate knowledge sharing and coordination with other EU-FP7 projects that covered similar activities during the same period. WP7-8 coordinated ERA-CLIM2 communication and outreach activities, including briefings to the European Commission and coordination with the EU-FP7 sister projects (CLIPC, EUCLEIA, QA4ECV, UERRA) that have been running in parallel with ERA-CLIM2, and aimed at building European capacity in climate change monitoring and in the establishment of climate changes services. WP9 included the ERA-CLIM2 project management and coordination activities.

ERA-CLIM2 plaid a key role in organizing the 5th International Conference on reanalysis (ICR5, Fig. 6), which was held in Rome in November 2017. At ICR5, many ERA-CLIM2 results were presented to the international community and reanalyses' users.



Figure 6. Conference media briefing at ICR5 (the 5th International Conference on reanalysis), held in Rome in November 2017. From left to right: Silke Zollinger (Copernicus ECMWF), Paul Poli (Météo-France), Roberto Buizza (ECMWF, ERA-CLIM2 coordinator), Michel Rixen (WCRP), Jean-Noël Thépaut (Copernicus ECMWF), and Simona Masina (CMCC).