Subseasonal deterministic prediction skill of low-level geopotential height affecting southern Africa
Loading...
Date
Authors
Engelbrecht, Christina Johanna
Phakula, Steven
Landman, Willem Adolf
Engelbrecht, Francois A.
Journal Title
Journal ISSN
Volume Title
Publisher
Universitätsverlag Potsdam
Abstract
The NCEP CFSv2 and ECMWF hindcasts are used to explore the deterministic subseasonal predictability of
the 850-hPa circulation of a large domain over the Atlantic and Indian Oceans that is relevant to the weather and climate of
the southern African region. For NCEP CFSv2, 12 years of hindcasts, starting on 1 January 1999 and initialized daily for four
ensemble members up to 31 December 2010 are verified against ERA-Interim reanalysis data. For ECMWF, 20 years of
hindcasts (1995–2014), initialized once a month for all the months of the year are employed in a parallel analysis to investigate the predictability of the 850-hPa circulation. The ensemble mean for 7-day moving averages is used to assess the
prediction skill for all the start dates in each month of the year, with a focus on the start dates in each month that are
representative of the week-3 and week-4 hindcasts. The correlation between the anomaly patterns over the study domain
shows skill over persistence up into the week-3 hindcasts for some months. The spatial distribution of the correlation
between the anomaly patterns show skill over persistence to notably reduce over the domain by week 3. A prominent area
where prediction skill survives the longest, occur over central South America and the adjacent Atlantic Ocean.
Description
Keywords
Climate prediction, Ensembles, Forecast verification/skill, Hindcasts, Operational forecasting, Seasonal forecasting
Sustainable Development Goals
Citation
Engelbrecht, Christien & Phakula, Steven & Landman, Willem & Engelbrecht, Francois. (2021). Subseasonal Deterministic Prediction Skill of Low-Level Geopotential Height Affecting Southern Africa. Weather and Forecasting. 36. 1-1. 10.1175/WAF-D-20-0008.1.