Potential benefit of extra radiosonde observations around the Chukchi Sea for the Alaskan short-range weather forecast

Abstract

In recent years, growing attentions have been paid to the potential benefit of extra observations over the data-sparse Arctic Ocean for weather forecasts. Here we also focus on such a case by targeting the inhabited land area, Alaska. During 2？18 August 2015, ship-borne radiosonde sounding observations were performed every 12 hour (except 6-hourly from 12:00 UTC 11 August to 00:00 UTC 14 August) around the Chukchi Sea. To assess the impact of those extra radiosonde observations, two sets of ensemble forecast experiments (CTLf and OSE_Af) were produced, which were respectively initialized by atmospheric reanalysis data without (CTL) and with (OSE_A) additional assimilation of those data. The tropospheric circulation fields are compared to verify their differences in forecast performance. While two forecasts have similar performance in the earlier spin-up period of the analysis-forecast cycle (from 4 to 7 August), their performance tends to diverge in the later period (from 11 to 18 August) due to the accumulated influence on the error reduction in OSE_Af. Among the improved forecasts in OSE_Af, two most outperformed forecasts, each initialized on 00:00 UTC 12 and 00:00 UTC 14, show a notable improvement in predicting the developing trough over Alaska on 16？17 August by suppressing the development of erroneous high anomalies in CTLf. Though the positive impact of single-point observations is limited in a space, our results suggest that enhanced radiosonde profile observations in the data-sparse polar ocean could be beneficial for the forecasts beyond the observational area.