Exploring the Propagation of the Madden-Julian Oscillation across the Maritime Continent Midshipman 1/C Casey R. Densmore Advisor: Professor Bradford S. Barrett, Oceanography Department, Advisor: CAPT Elizabeth R. Sanabia, Oceanography Department External Collaborator: Assistant Professor Pallav Ray, Florida Institute of Technology Trident Research Objective: Better understand atmospheric conditions which favor MJO eastward propagation through the Maritime Continent Introduction Data and Methods Result: Background Atmospheric States Classifying MJO Events Primary Result 1: Increased moisture and lower pressures are potential precursors for MJO propagation regardless of initial strength. A more intense lower The Madden-Julian Oscillation (MJO, Madden and Julian 1971) is the tropospheric moisture foot east of the active envelope also indicates likely MJO propagation for initially active events. These atmospheric anomalies may provide a • The RMM Index was used to categorize MJO events based leading intraseasonal mode of atmospheric variability. The MJO predictive capability to detect changes in the MJO prior to those changes projecting on the RMM index. consists of a broad circulation cell approximately 10,000 km in on location and strength • Categorized as one of four cases based on MJO horizontal extent (Fig. 1; Madden and Julian 1994) that propagates Specific Humidity activity on entrance to and exit from the Maritime eastward along the equator circumnavigating the globe in the • AA MJO events are significantly more moist throughout the troposphere in the active tropics (Zhang 2005) over a period of approximately 30 to 60 days. Continent (Fig. 4, matching Fig. 2). • Eastward propagation required for 65% of days over envelope than AI events from days -8 to +8 (Figs. 7C,F,I) Figure 7: Specific humidity (color- • Positive low-level humidity (moisture foot) extends eastward over the MC at day 0 filled) and geopotential height MJO circulation (Figs. 1,2) includes: the MC • Signature is significantly more moist for AA events than AI events (Fig. 7F) (unfilled contours, positive • Upward motion (“active envelope”) • Analysis was repeated for: represented by solid lines) • Above-normal thunderstorm activity due to upward • OLR MJO (OMI) index to confirm results (not shown) • No moisture foot signature present for MJO events over the IO or WP (not shown) standard anomaly contours • Summer with BSISO index to better capture different • Positive humidity anomalies facilitate enhanced convective development which would averaged from 15 °S to 15 °N as vertical motion, causing above-average precipitation the MJO active envelope crosses and wind summer thunderstorm activity enhance MJO propagation the western boundary of the MC • Downward motion (“suppressed envelope”) • MJO events over the Indian Ocean (phases 2 and 3) (100 °E). Anomalies are shown for and western Pacific Ocean (phases 6 and 7) to AA (A, D, G) and AI (B, E, H) MJO • Below-normal thunderstorm activity due to downward Geopotential Heights events, as well as the differences Figure 1: Atmospheric circulation and resultant convective compare to MC characteristics. vertical motion, causing below average precipitation Figure 4: MJO events over the Maritime Continent • Negative low-level and positive upper-level geopotential height anomalies larger for AA between the two (C, F, I) at -8 (A- activity typical of the MJO. 2014 NOAA graphic by Fiona • Analysis was modified to study relationships between MJO C), 0 (D-F), and 8 (G-I) days relative and wind Martin. from 1980-2015 classified as one of four cases by events than AI events by day 0 (Fig. 7F), amplitude changes over MC and phase of the QBO phase 4 and 5 RMM amplitudes: Active-Active (AA, to MJO active envelope entrance Quantifying the MJO • Mean amplitude changes over the MC for all AA and A), Active-Inactive (AI, B), Inactive-Active (IA, C), • Corresponds to lower surface pressure, and faster upward vertical motion for AA events to the MC. The Wheeler-Hendon Realtime Multivariate (RMM) MJO index (Wheeler and Hendon 2004; WH04) uses wind and AI MJO events were compared by QBO phase and Inactive-Inactive (II, D). Black lines denote • Larger geopotential height anomalies enhance circulation in the active envelope thereby average amplitudes for all MJO events of each case. outgoing longwave radiation (OLR) to quantify MJO (Fig. 4): strengthening the overall MJO circulation. • Geographic location: one of eight phases • Strength: Active or inactive based on strength of thunderstorm and wind anomalies. Calculating Anomalies of Background Atmospheric States • Daily values of specific humidity (q) and geopotential heights Result: MJO-QBO Relationship Region of Interest (ɸ) from the ERA-Interim Reanalysis were compared among QBO and Lower-Stratospheric Stability QBO and MJO Amplitude Change by Season A B the four MJO cases at 0000 and 1200 UTC from 1980-2015. q − One region particularly susceptible to strong MJO- • i, j j Primary Result 2: During winter, QBO winds drive stability changes in the Primary Result 3: MJO propagation exhibits a statistically significant relationship Standard anomalies were calculated following Eq. 1, where: q = (1) driven thunderstorms is the Maritime Continent (MC, • q is the parameter analyzed, μ is the climatological stdanom i lower stratosphere that correspond to stronger MJO events propagating with QBO that reverses with season. In northern hemisphere winter, MJO events Inness and Slingo 2006), a region of island nations and monthly mean from 1980-2015, and σ is the standard j across the MC. The impact of these QBO-driven stratospheric zonal winds strengthen over the MC when mid-stratospheric winds are easterly and weaken archipelagos in Southeastern Asia including Indonesia. deviation of the monthly mean on lower-stratospheric stability and MJO depends on the altitude of those when those winds are westerly. During northern hemisphere spring and summer, • Statistical significance was tested at the 95% confidence level winds, as well as their direction and speed. MJO and BSISO events weaken more when mid-stratospheric winds are easterly. MJO Propagation through the MC During boreal winter, zonal wind shear is in thermal wind balance with The relationship between QBO and MJO propagation over the MC is significant and As the ascending branch of the MJO envelope reaches Quantifying the QBO with an EOF analysis stratospheric temperature anomalies. reverses with season. the MC, the convective signal either: EOF analysis of global stratospheric zonal winds 2 During winter, temperature D • Moves eastward through the MC and reaches the 푑푢 푅 δ 푇 푑푢 C from 1980-2017 (data from ERA-Interim, Dee et. = − ∝ 푇′ (2) anomalies (T’) are proportional western Pacific Ocean as active in a propagating 2 푑푧 al, 2011) 푑푧 퐻훽 δ푦 to zonal wind shear (du/dz) (Fig. event (Figs. 2A, C) • Averaged meridionally from 10°S to 10°N • Fails to reach the western Pacific Ocean as active in a 8, Eq. 2; (Holton and Hakim, • Bounded vertically by 10 hPa and 100 hPa 2013) ). non-propagating event (Figs. 2B,D; Li and Feng 2015) • Most variance is captured by the first two PCs in the 2-3 year time period (Fig. 5) MJO propagation through the MC, the focus of this Changes in zonal winds with Trident research, is challenging to predict, and is height therefore create warm currently an area of active research by the and cold temperature anomalies Figure 2: Zonal and vertical wind anomalies (red arrows) and resultant meteorological community. in the stratosphere. convective activity in all four cases of MJO events (AA, AI, IA, and II). Left and right blue shading represents the Indian and Western Pacific Oceans, respectively, and green shading represents the MC. Figure 5: Variance density spectra of the first These temperature changes three principal components. drive stability changes which can Quasi-Biennial Oscillation (QBO) and MJO Phase space diagram (Fig. 6) shows QBO wind speed either enhance or weaken high- Figure 9: Mean amplitude change difference from climatological (1980-2017) average for initially active MJO and height: Figure 8: Zonal wind shear values (thin lines) and altitude thunderstorm activity events, binned by QBO phase pair and seasonal subset. Boreal summer (JJA) analysis is repeated using MJO RMM Other atmospheric oscillations can project onto temperature anomalies (thick lines) associated with each amplitudes and BSISO amplitudes. • Angle around center corresponds to altitude, weather patterns with shorter timescales, including of the four QBO phase pairs. associated with the MJO. direction of winds the MJO. Previous studies have linked MJO Winter (December-February) • Radius from center corresponds to wind speed amplitude to the polarity of one such interannual • QBOEM: MJO strengthens over the MC • QBO winds can be categorized by zonal wind Winter QBO-MJO Interaction oscillation: the Quasi-Biennial Oscillation (QBO; e.g. • Easterly zonal wind shear cools the lower stratosphere, lowering lower direction and height as one of four phase pairs: Son et al. 2017). stratospheric stability and enhancing deep convection associated with the • QBOWL: Westerly lower-stratospheric winds MJO (Fig. 10) The QBO consists of alternating centers of easterly Figure 6: EOF phase space diagram of QBO1 and • QBOEM: Easterly mid-stratospheric winds QBO2 from 1980-2017. Eight phases (separated by • QBOWM: MJO weakens over the MC and westerly winds descending through the • QBOEL: Easterly lower-stratospheric winds dashed lines) divide the