In order to explore the sensitivity of the climate impact of volcanic eruptions to eruption season and latitude, we simulate volcanic eruptions at different latitudes and in different seasons with the Max Planck Institute Earth System Model (MPI-ESM). We use the same configuration of the MPI-ESM model as used for the historical simulation of CMIP6. An initial run is performed firstly (PINArst). Then we perform 23 and 10 control runs without any volcanic eruption (PINAref and PINAwRef). Two groups of three different latitudinal volcanic eruptions in boreal summer and winter are simulated. We perform 10-member simulations for each eruption case. 9 Tg of total sulfur injection magnitude is prescribed. The eruption latitudes are set to be 0° for the equatorial eruptions (PINAeq and PINAwEQ) and 30° N and 30° S for the northern and southern hemispheric eruptions (PINAnh, PINAwNH, PINAsh and PINAwNH), respectively. For the summer eruptions, the date is set to be the same as the 1991 Pinatubo eruption on June 15, 1991; for the winter eruptions, the date is set to be December 15, 1991.

Hindcast simulation has been performed for reconstruction of meso-scale weather systems, especially of intense storms by using the global atmospheric model ECHAM6 in its version 6.1.00 (CMIP5), the spectral nudging technique and NCEP reanalysis as forcing data. Horizonatal resolution: (768x384) gridpoints (T255L95), vertical resolution: 95 levels. Time period: 1948 - 2015 april. The output from the model run: /work/gg0301/g260070/ECHAM6/echam-6.1.00/experiments/echam6_t255l95_sn_ncep1/scripts on the HLRE-2 Blizzard and the Original-Outputs of the 8 ECHAM6-streams were named and differed by their suffixes: echam6_t255l95_sn_ncep1_199212.01_co2 echam6_t255l95_sn_ncep1_199212.01_echam echam6_t255l95_sn_ncep1_199212.01_jsbach echam6_t255l95_sn_ncep1_199212.01_land echam6_t255l95_sn_ncep1_199212.01_nudg echam6_t255l95_sn_ncep1_199212.01_surf echam6_t255l95_sn_ncep1_199212.01_veg echam6_t255l95_sn_ncep1_199212.01_vphysc

RCM forcing data from the first realisation of the CMIP5 experiment rcp85. The experiment covers the period 2006-2100. Realisation is started end of 2005 from the respective realisation of the historical experiment.

RCM forcing data from three realisations of the CMIP5 experiment historical. The historical experiment covers the period 1850-2005. Forcing data are available from 1949 to 2005 only. The realisations differ with respect of their branch times of the pre-industrial control run (piControl): historical_r1i1p1 started end of 1879, historical_r2i1p1 started end of 1979 and historical_r3i1p1 started end of 1920.

RCM forcing data from three realisations of the CMIP5 future scenario experiment rcp26. The experiment covers the period 2006-2100. The members are started end of 2005 from the respective members of the historical experiment.

RCM forcing data from three realisations of the CMIP5 experiment rcp45. The experiment covers the period 2006-2100. The members are started end of 2005 from the respective members of the historical experiment.

RCM forcing data from the first realisation of the CMIP5 experiment amip. The experiment covers the years 1979 to 2008 with a three year initialized spinup phase. (Realisations differ by a small disturbance of the initial state only, and therefore by their restart files only)

RCM forcing data from the first realisation of the CMIP5 experiment rcp45. The experiment covers the period 2006-2100. Realisation is started end of 2005 from the respective realisation of the historical experiment.

RCM forcing data from three realisations of the CMIP5 experiment decadal2005. The decadal2005 experiment covers the years 2006-2035. The members start from different states in late 2005 (1 day apart) of an assimilated historical run (realisation 1).

RCM forcing data from three realisations of the CMIP5 experiment decadal1990. The decadal1990 experiment covers the years 1991-2000. The members start from different states in late 1990 (1 day apart) of an assimilated historical run (realisation 1).