Intraseasonal Variability in Aquaplanet Configuration of Community Atmosphere Model

Abstract

The Intraseasonal Oscillation (ISO) plays an important role to modulate deep convective activity in the tropical region. In this thesis, I aim to understand the role of land and warm oceans in ISO, using a general circulation model. For this, I conduct a series of experiments in the Community Atmosphere Model (CAM) with various idealized and realistic surface boundary conditions to study tropical ISO. To investigate the influence of tropical sea surface temperature (SST) on ISO and convectively coupled equatorial waves in the global atmosphere, I conduct experiments with idealized, zonally symmetric SST profiles having different widths of warm ocean centered at the equator. I use the model in its basic “Aquaplanet” configuration, with the sun at the equator, i.e. perpetual spring equinox forcing; with idealized zonally symmetric SST, the aquaplanet model produces a double Intertropical Convergence Zone (ITCZ) on either side of the equator, and an eastward propagating Madden Julian oscillation (MJO)like mode with variance at intraseasonal (30 to 96 day) periods and zonal wavenumber one. In the experiment with the narrowest meridional width of warm SST, the variance of moist convective activity lies predominantly in equatorially trapped Kelvin wave band. As the width of the warm equatorial SST is increased, the eastward propagating speed of the MJO-like signal decreases; for the broadest SST profile (warm SST covering 20 degrees of latitude), the speed of the model MJO is about 5.5 m s−1, close to the observed speed. This is because the latitudinal extent of warm SST is comparable to the equatorial Rossby radius, and the model produces off equatorial Rossby waves of sufficient strength to interact with the Kelvin wave and slow down the MJO-like mode. The model also generates westward propagating waves with intraseasonal periods and zonal wavenumber 1–3; the structure of these signals, which extend well into the mid-latitudes, projects onto equatorially trapped Rossby waves with meridional mode numbers 1, 3 and 5, associated with convection that is symmetric about the equator. In addition, the model generates 30–80 day westward moving signals with zonal wavenumber 4–7, particularly in the experiment with a narrow region of warm SST. Although these waves are seen in the wavenumber-frequency spectra in the equatorial region, they have the largest amplitude in the middle and high latitudes. Thus, our study shows that wider, meridionally symmetric SST profiles support a strong MJO-like eastward propagation, and even in an aquaplanet setting, westward propagating Rossby waves comprise a large portion of tropical intraseasonal variability. In the observations (ERA-Interim daily reanalysis), the MJO signal lies in the range of zonal wavenumbers 1 to 5. The variance of MJO at higher wavenumbers (2–5) is absent in the aquaplanet model. For this, I design model experiments in order to study how model MJO responds to the introduction of continents in the presence of zonally symmetric SST, and a realistic SST distribution with the Indo-Pacific warm pool and cool SST in the eastern Pacific. As before, the model is in the aquaplanet-like configuration, to eliminate the effects of seasonality. Model results are compared with 21 years (1995–2015) ERA-Interim reanalysis data and analyzed in terms of the moist static energy (MSE) budget to study the growth and propagation of MJO. When I introduce continents with realistic orography and interactive surface temperature, soil moisture, and albedo, the variance of model MJO is reduced due to weaker boundary layer moisture convergence. However,MJO variance extends to higher wavenumbers. With prescribed climatological January SST boundary condition in the presence of continents, the variance of model MJO is enhanced by a factor of 2–3, and it is distributed across zonal wavenumbers 1 to 5, in closer agreement with observations. Thus, I find that the presence of land by itself is not enough to produce realistic MJO in CAM, but realistic SST distribution is also necessary to simulate MJO with improved spacetime characteristics. Both in simulations and ERA-Interim data, column-integrated longwave radiation plays a key role in the growth of MSE anomaly associated with MJO; in general, meridional and vertical advection of MSE both acts to promote eastward movement of MJO. In the model experiments, meridional advection of low-level MSE anomaly is most significant in the vicinity of the ITCZ. This indicates that the physical processes which determine the location of (single or double) ITCZ are linked to MJO dynamics.

The westward propagating “quasi-biweekly” oscillation (QBWO) with 10–25 day period is an important intraseasonal mode of the Asian summer monsoon, yet very few model studies focus on this mode. I study QBWO in the northern and southern tropics in the model and compare it with ERA-Interim reanalysis data. The pure aquaplanet model produces a double Intertropical Convergence Zone (ITCZ), winds that are predominantly zonal, and weak quasi-biweekly variance. When continents are introduced in the model with zonally symmetric SST, the northern ITCZ, as well as quasi-biweekly variance between 10◦N to 24◦N are strengthened in the Pacific Ocean, bringing model results closer to observations. In the model with continents, the QBWO signal dwells inside the mean envelope of high atmospheric moisture, or total precipitable water (TPW), in agreement with observations. However, in the presence of zonally symmetric SST, the model fails to simulate sufficiently high precipitable water in the region extending from the north Indian Ocean to East Asia, resulting in very weak QBWO variance. When the model includes continents and realistic (January) SST boundary conditions, the spatial structure of both TPW and QBWO variance becomes more realistic. I study the mechanisms of propagation and maintenance of the quasi-biweekly mode using vorticity budget and moist static energy (MSE) budget analysis. Advection due to the effect is responsible for the northwestward propagation of QBWO vorticity, while the propagation of column MSE anomaly is mainly due to horizontal advection. Surface turbulent heat fluxes and vertical MSE advection are the dominant contributors to the growth and maintenance of column MSE anomaly in observations and model respectively. Surface heat flux makes a significant contribution to the growth of quasi-biweekly MSE anomaly in the presence of land, in association with the enhanced meridional wind, and vortical structures that resemble moist Rossby waves with a wavelength of about 4000 kilometers.