Maurice Blackmon Symposium October 29 2007 Intraseasonal LowFrequency

Maurice Blackmon Symposium October 29, 2007 Intraseasonal Low-Frequency Variability: At the Interface Between Weather and Climate Randall M. Dole NOAA Earth System Research Laboratory Physical Sciences Division Maurice (shortly after leaving NOAA)

Meeting Maurice NCAR 1978 ASP Summer Colloquium The General Circulation: Theory, Modeling and Observations Coordinator: Maurice Blackmon Principal Lecturers: Bob Gall Isaac Held Brian Hoskins Dick Lindzen Kevin Trenberth Mike Wallace Additional seminar speakers: Steve Ashe, Bob Dickinson, Gabriel Lau, Ed Lorenz, Rol Madden, Jerry Mahlman, Mike Mc. Intyre, Jan and Julia Paegle, Dave Salstein, K. K. Tung, Harry van Loon. Jim Mc. Williams, among others, was also quite involved in certain debates.

Colloquium Experience: A Few Reflections I was one of 14 students participating in this Colloquium. Most were supervised by their faculty advisors, but I came from MIT as a “free agent”. Fortunately, Maurice agreed to serve as my summer science advisor, because this colloquium, unlike previous ones, had as a major component student research projects. This was a fantastic experience! Also: 1)This colloquium occurred during a transformative time in general circulation research. As noted by Mike Wallace at the time, over half of the presentations were on topics beyond the zonally-averaged circulation. 2) Exciting new research was appearing on the three-dimensional structure of the general circulation and variability on different time and space scales, led by Maurice, Gabriel Lau, Mike Wallace and colleagues. It opened up whole new ways to attack certain key questions. 3)The colloquium emphasized the vital links between observations and theory. • For my summer project, “The Objective Representation of Blocking Patterns”, I applied some techniques inspired by Maurice’s approaches to a specific phenomenological problem. At that time, blocking events were perhaps the best recognized synoptic example of what we now call intraseasonal low-frequency variability.

Some initial questions on Low Frequency Variability (LFV) • What is the geographical distribution of LFV? • Are there characteristic flow patterns associated LFV? • What were the synoptic manifestations of LFV? • Was LFV associated uniquely with blocking, or a broader class of recurrent phenomena that had time scales longer than typical synoptic periods? For example, were there “anti-blocks” as well as blocking events? If so, could such features be identified within a more general and systematic framework? • Could relationships between LFV and changes in synoptic-scale storm variability be rigorously established? • Could fundamental dynamical mechanisms for LFV be identified?

Intraseasonal LFV Geographic Distribution: Variability in wintertime NH 500 mb height fields (Blackmon 1976) H rms variability (m) H H H Low-pass variability (m) • Blackmon low-pass (LP) variability retains periods ~ 10 -90 days. Time-mean heights (m) J • Total rms variability of 500 mb height fields is dominated by low-pass fluctuations. L J • Low-pass fields show three primary “centers of action”: North Pacific, North Atlantic, and far northern Russia. First two centers are located downstream of oceanic jet maxima.

Relationships to storm tracks (Blackmon et al. 1977 and other studies) Bandpass variability sea-level pressure (mb) H HH H Lowpass variability sea-level pressure (mb) • Blackmon bandpass (BP) filtering procedure retains periods ~ 2. 5 -6 days. This variability is closely related to storm tracks. Two dominant oceanic maxima. • LP variability in SLP also shows three dominant maxima, approximately co-located with the 500 mb height variability maxima (suggests a substantial barotropic component). • Nature of LP fluctuations not as clearly revealed as for BP fluctuations. Locations of oceanic LP centers downstream from BP maxima suggests they may be partly associated with mature or decaying baroclinic storms.

Vertical (Blackmon structures et al. 1979) Shapes of dominant vertical modes for selected regions together with map showing correlation between 1000 mb and 500 mb height variability. • Typical vertical structures near LP maxima have peaks near tropopause, and relatively high positive correlation between 1000 mb and 500 mb heights, again suggesting a relatively more barotropic structure. Some caution needed in this interpretation.

Horizontal structures of LFV - A new (old) paradigm: Teleconnections Teleconnection example: The Pacific - North American (PNA) pattern Schematic from Wallace and Gutzler (1981) illustrating several dominant teleconnection patterns.

Time evolution - synoptic manifestations _ Composite evolution of positive phase of PNA pattern in 300 mb height field (anomalies shaded). Day 0 is defined by a threshold crossing procedure (23 cases from Dole and Black 1990). _ + _ Common characteristics: • Development occurs rapidly (over several days). • Pacific cases are preceded by a zonally-elongated perturbation upstream that propagates into the main development region. • Following development, there is energy dispersion downstream. _ + _ _ + + _ _

Some other observational characteristics of intraseasonal LFV • There is a surprising degree of linearity of LFV patterns with respect to the time-mean flow, e. g. , as evidenced by teleconnections, (modest deviations from linearity occur at large eddy amplitudes). The high degree of linearity is extremely useful for many observational and theoretical analyses. • For extratropical intraseasonal LFV, there is no clear evidence for periodicity or a preferred duration (red noise-like behavior, again with modest deviations). • A profusion of teleconnection patterns have been identified; however, the total number of patterns required to describe extratropical LFV is still relatively small - perhaps fewer than we might have originally anticipated. The patterns themselves should not be interpreted as fixed physical entities, but rather are useful to the extent that they reveal certain dominant or frequently recurrent physical and dynamical processes. • Numerous studies have shown a close connection between LFV and extratropical storm tracks. This is a two-way street: the LFV alter the storm tracks and eddy life cycles, while the eddies produce feedbacks which alter the evolution of the LFV. The ability to predict LFV and interactions with the higher-frequency variations is at the heart of most efforts to improve extended-range weather forecasts.

Progress toward understanding LFV Rossby wave propagation on a sphere (e. g. , Hoskins et al. 1977). • Provides a fundamental basis for understanding energy dispersion away from localized sources; for example, emanation of a Rossby wavetrain downstream from Pacific during PNA development. • Not clear that a theory developed for a zonally-symmetric basic state could explain the localized maxima in LFV described by Blackmon (1976). To do this would require zonally asymmetric wave sources. These existed, but were they enough? Vorticity Streamfunction, ’ Response of a barotropic, super-rotating atmosphere to forcing by a circular mountain at 30 N (from Grose and Hoskins, 1979).

Barotropic Instability on a wavy basic state Evolution (at 4 day intervals) of the streamfunction of the most unstable barotropic mode in a zonally-varying basic state similar to the NH winter time circulation (from Simmons et al. 1983). Theory (Hoskins et al. 1983, Simmons et al. 1983, SWB) can help explain tendency for maximum LFV variability downstream of jet stream maxima, but barotropic model also has limitations.

Comments on LFV mechanisms • Many processes can produce LFV, and undoubtedly do. These include i) slow variations in forcing or in the background state itself (e. g. , associated with the annual cycle), ii) transient growth by either barotropic or baroclinic processes from suitably configured initial perturbations, and iii) interactions with synoptic-scale eddies. Neat stuff, and is likely to be discussed in other presentations here. • What the previous studies established is the fundamental importance of Rossby wave dispersion as well as of climatological stationary waves (i. e. , wavy basic states) in accounting for important observed characteristics of LFV as described by Maurice and other colleagues, including the geographical distribution, typical horizontal and vertical structures, and time evolution of variability on the 10 -90 day time scale.

Science challenges for future LFV research • Modeling and prediction of organized tropical convection (MJO, ISO …) • Tropical-extratropical interactions (another two-way street). • Stratosphere-troposphere interactions, often prominent with high latitude modes. • Possible impacts of coupled processes, e. g. , ocean-atmosphere interactions, or atmosphere-land surface interactions for summer events such as droughts. • Connections with longer time scale phenomena, e. g. , ENSO and beyond, as well implications for climate change. • Interactions with higher-frequency “weather” processes, including moisture transports. • Implications of LFV for predicting weather and climate extremes, as well as for anticipating how these events may change in future climates.