5:15 PM - 7:15 PM
[ACG52-P13] Evaluation of temperature forecasts in the marginal ice zone of the Chukchi Sea during the R/V Mirai Arctic expedition in November 2018
Keywords:marginal ice zone, numerical weather prediction, Arctic observations
Observations in the Arctic region have become increasingly important to accurately monitor these rapid changes. Observations are also important to enhance our understanding of the interactions among atmosphere, ocean, and sea ice on various timescales, which elaborate the coupled atmosphere-ocean–sea ice models. Because sea ice loss opens opportunities for non-icebreakers to use the Arctic shipping routes, accurate atmospheric, ocean, wave, and sea ice predictions are important to navigate Arctic shipping safely. In November 2018, R/V Mirai conducted an Arctic expedition in the refreezing Chukchi Sea (MR18-05C, JAMSTEC, 2018). An extensive observation campaign was conducted near the marginal ice zone (MIZ) during the expedition. This study evaluated the ECMWF operational forecasts using observational datasets from the Arctic expedition.
The temperature differences between the ECMWF analysis and surface and lower-level troposphere observations were positive near the MIZ from 9 to 15 November 2018 (Fig. 1a). The temperature errors were also positive in the 24-hour, 48-hour, and 72-hour forecasts, peaking on 13 November 2018 (Fig. 1b-d). All ensemble members showed similar positive errors, suggesting that the ECMWF prediction failed to predict the temperature decrease during this period. In the forecasts initialized from November 9 to 12, the sea ice concentration (SIC) in the model was lower than that estimated by the AMSR2 retrieval in the MIZ of the eastern Chukchi Sea (72º–76ºN, 180º–190ºE). The surface heat flux peaked locally at this location in the model, suggesting that heat and moisture were being transferred from the ocean to the atmosphere in the region. A wind error depicted a cyclonic circulation centered on the area of negative SIC difference. From the ERA5, the latent heating was estimated at approximately 4K/day at 900 hPa, and potential vorticity peaked at 950 hPa. These results indicate that the heat and moisture supply from the ocean due to the lesser SIC in the MIZ contributed to the latent heating over the region. The resulting cyclonic circulation would lead to the positive temperature errors observed at the location of R/V Mirai during this period.
The temperature differences between the ECMWF analysis and surface and lower-level troposphere observations were positive near the MIZ from 9 to 15 November 2018 (Fig. 1a). The temperature errors were also positive in the 24-hour, 48-hour, and 72-hour forecasts, peaking on 13 November 2018 (Fig. 1b-d). All ensemble members showed similar positive errors, suggesting that the ECMWF prediction failed to predict the temperature decrease during this period. In the forecasts initialized from November 9 to 12, the sea ice concentration (SIC) in the model was lower than that estimated by the AMSR2 retrieval in the MIZ of the eastern Chukchi Sea (72º–76ºN, 180º–190ºE). The surface heat flux peaked locally at this location in the model, suggesting that heat and moisture were being transferred from the ocean to the atmosphere in the region. A wind error depicted a cyclonic circulation centered on the area of negative SIC difference. From the ERA5, the latent heating was estimated at approximately 4K/day at 900 hPa, and potential vorticity peaked at 950 hPa. These results indicate that the heat and moisture supply from the ocean due to the lesser SIC in the MIZ contributed to the latent heating over the region. The resulting cyclonic circulation would lead to the positive temperature errors observed at the location of R/V Mirai during this period.