Fig. 2: Floods and meteorological conditions in West Sumatra in 2017.
From: Social-media and newspaper reports reveal large-scale meteorological drivers of floods on Sumatra
Time series of flood and meteorological conditions during 2017 in West Sumatra. a 3-hourly Twitter data (gray line) and 3-day running average of Twitter data (black line); colored bars indicate floods identified in Twitter (red), paper (blue) and Indonesian National Board for Disaster Management (BNPB) (magenta) databases in 2017, (b) daily precipitation accumulation [mm] averaged over West Sumatra region from TRMM (black line), averaged over rain-gauge locations only from TRMM (blue dashed line) and maximum value of the three rain-gauges (red line), (c) 5-day precipitation accumulation [mm] averaged over the West Sumatra region in 2017 and Madden Julian Oscillation (MJO) phase Real-time Multivariate Madden Julian Oscilation index (RMM) phases 2–5 for RMM amplitude above 1 only, see legend), (d) convectively coupled Kelvin wave (CCKW) precipitation rate anomaly [mm/day] averaged over the the West Sumatra region and the MJO phase (RMM phases 6–1 for RMM amplitude above 1 only, see legend). Thick black lines in d indicate the robust CCKWs defined as the waves with a well defined trajectory20 at least between longitudes 80E and 110E. The yellow shading in all panels indicates major flooding periods for which identification from Twitter, Papers and BNPB agree.
