Back to Home.

The plots below show:

- the time series of Arctic sea-ice area (Fig. 1) and volume (Fig. 2) in March and September
- the latitudinal transect of northward Atlantic ocean heat transport averaged over 2000-2014 (Fig. 3)
- the latitudinal transect of northward Pacific ocean heat transport averaged over 2000-2014 (Fig. 4)
- the time series of northward Atlantic ocean heat transport at 60N and 70N (Fig. 5)
- the detrended March Arctic sea-ice area against the detrended annual mean Atlantic ocean heat transport at 70N for all points between 1950 and 2014 for some models (Fig. 6)
- the detrended March Arctic sea-ice volume against the detrended annual mean Atlantic ocean heat transport at 70N for all points between 1950 and 2014 for some models (Fig. 7)

from the CMIP6 historical and SSP1-2.6/SSP5-8.5 runs (data kindly downloaded by Klaus Zimmermann and Michael Sahlin at SMHI). Also shown are the multi-model means (thick dashed black curve) and observations/reanalysis (thick solid black curve).

Fig. 1a: Time series of Arctic sea-ice area in (a) March and (b) September for historical runs. The thick dashed black curve is the multi-model mean and the thick solid black curve is the OSI SAF satellite dataset. The number in brackets indicates the number of members to compute the mean sea-ice area for each model (number of models for the multi-model mean).

Fig. 1b: Time series of Arctic sea-ice area in (a) March and (b) September for SSP1-2.6. The thick dashed black curve is the multi-model mean. The number in brackets indicates the number of members to compute the mean sea-ice area for each model (number of models for the multi-model mean).

Fig. 1c: Time series of Arctic sea-ice area in (a) March and (b) September for SSP5-8.5. The thick dashed black curve is the multi-model mean. The number in brackets indicates the number of members to compute the mean sea-ice area for each model (number of models for the multi-model mean).

Fig. 2a: Time series of Arctic sea-ice volume in (a) March and (b) September for historical runs. The thick dashed black curve is the multi-model mean and the thick solid black curve is the PIOMAS reanalysis dataset. The number in brackets indicates the number of members to compute the mean sea-ice volume for each model (number of models for the multi-model mean).

Fig. 2b: Time series of Arctic sea-ice volume in (a) March and (b) September for SSP1-2.6. The thick dashed black curve is the multi-model mean. The number in brackets indicates the number of members to compute the mean sea-ice volume for each model (number of models for the multi-model mean).

Fig. 2c: Time series of Arctic sea-ice volume in (a) March and (b) September for SSP5-8.5. The thick dashed black curve is the multi-model mean. The number in brackets indicates the number of members to compute the mean sea-ice volume for each model (number of models for the multi-model mean).

Fig. 3: Latitudinal transect of northward Atlantic-Arctic ocean heat transport (OHT) averaged over 2000-2014. The thick black solid line is the OHT estimate from Trenberth et al. (2019) and the grey shading around it represents the temporal 2-standard deviation range. The two black dots show the hydrographic measurements at 26N (RAPID-MOCHA, 2004-2018) and 57N (OSNAP, 2014-2016) with their associated temporal standard deviation represented as error bars. The number in brackets indicates the number of members to compute the mean OHT for each model.

Fig. 4: Latitudinal transect of northward Indo-Pacific ocean heat transport (OHT) averaged over 2000-2014. The thick black solid line is the OHT estimate from Trenberth et al. (2019) and the grey shading around it represents the temporal 2-standard deviation range. The number in brackets indicates the number of members to compute the mean OHT for each model.

Fig. 5: Time series of northward Atlantic ocean heat transport (OHT) at (a) 60N and (b) 70N. The thick black solid line is the OHT estimate from Trenberth et al. (2019). The number in brackets indicates the number of members to compute the mean OHT for each model.

Fig. 6: Scatter plots of detrended March Arctic sea-ice area against detrended annual mean Atlantic ocean heat transport at 70N for all points between 1950 and 2014. We show results only from the 5 best CMIP6 models in terms of historical Arctic SIA and Atlantic OHT, as well as observations/reanalysis in the bottom right panel (OSI SAF for sea-ice area and Trenberth et al. (2019) estimate for ocean heat transport). The correlation coefficients *R* and regression slopes *a* (unit: 1000 km^2 per TW) are shown in the upper right corner of each panel for each model.

Fig. 7: Scatter plots of detrended March Arctic sea-ice volume against detrended annual mean Atlantic ocean heat transport for all points between 1950 and 2014. We show results only from the 5 best CMIP6 models in terms of historical Arctic SIA and Atlantic OHT, as well as reanalysis in the bottom right panel (PIOMAS for sea-ice volume and Trenberth et al. (2019) estimate for ocean heat transport). The correlation coefficients *R* and regression slopes *a* (unit: km^3 per TW) are shown in the upper right corner of each panel for each model.

**Refinement of SSP1-2.6:**

Fig. 8: Time series of Arctic sea-ice area in (a) March and (b) September for SSP1-2.6 and for different selection criteria.

Fig. 9: Time series of Arctic sea-ice volume in (a) March and (b) September for SSP1-2.6 and for different selection criteria.

**Refinement of SSP5-8.5:**

Fig. 10: Time series of Arctic sea-ice area in (a) March and (b) September for SSP5-8.5 and for different selection criteria.

Fig. 11: Time series of Arctic sea-ice volume in (a) March and (b) September for SSP5-8.5 and for different selection criteria.

Important note: These figures were produced by D. Docquier and constitute preliminary results (not published in papers).