An assessment of MJO Impacts on U.S. precipitation: From intraseasonal predictability to decadal changes

Intraseasonal predictability is a challenging forecasting issue, as it lies between the initial-value problem of short-term forecasts and the boundary condition problem of seasonal forecasts. A known driver of intraseasonal forecasting is the Madden-Julian Oscillation (MJO). The MJO has the ability to impact U.S. weather on a regional scale with lead-times from days to weeks. It is therefore of value to forecasters to assess MJO-related precipitation changes in observations and models. To address this, we performed a composite analysis of precipitation as a function of MJO phase for the GEFSv12 model out to lead times of 35-days for the winter season (JFM). Comparison is made to two observation datasets; the first is CPC Global Unified Gauge-Based Analysis of Daily Precipitation and the second is ECMWF Reanalysis v5 (ERA5). Since the climate has been warming over the last several decades, a gridded-regression analysis is also shown for the ERA5 and CPC data, for which MJO composites are available from 1974 and 1979 respectively. The results of this work preliminarily addresses two questions: 1) How well does the GEFSv12 predict MJO-related precipitation changes on intraseasonal timescales and 2) Given that the climate is warming, have the MJO precipitation composites changed on decadal timescales? The latter question is of special importance to forecasters as it informs if previously-established MJO effects are still relevant.