Images of golden sandy beaches stretching magnificently along the sun-soaked Pacific were immediately crushed on arrival at the Best Western hotel stationed within the urban sprawl, in this case named Long Beach, that clings relentlessly to the entire Los Angeles basin. Although locals and meteorologists alike were fascinated by the anomalous down-pours that consumed the AMS meeting, this did not really add to the rather characterless location nor help to improve a slightly disappointing conference which boasted around 10 concurrent sessions within the impressive Long Beach Convention Center. Here follows a summary of talks that interested me:
John Bates talked on Modes of Tropical water vapour variability. The 1%/yr increase in lower stratospheric water vapour was discussed and it was postulated that changes in frequency of tropopause overshoot may be important in additional to the mechanism involving Methane changes. The surprising decadal changes in the radiation budget (e.g. Wielicki et al. 2002) were inconsistent with the lack of changes in HIRS free tropospheric humidity [I would note that the lack of changes in clear-sky OLR in the radiation budget instruments IS consistent with the HIRS humidity changes and make the cloudy portion of the radiation changes all the more interesting]. However, the high frequency variability in both HIRS and the radiation budget were greater than models and, in Bates' opinion, more interesting than the decadal changes. It was argued that interactions of Rosby wave activity in the E. Pacific during cold events interact with the subtropics in allowing re-hydration of the upper troposphere by increased vertical fluxes through transient eddies and that these are important in determining high frequency variability in the sub-tropical free-tropospheric water vapour.
Mark McCarthy, among other things, showed that the subtropical drying and tropical moistening trends, used by Bates and others as evidence of an intensification of the hydrological cycle, in fact appear unrelated to each other as most of the the subtropical drying occurred at a separate time to the tropical moistening.
John Christy argued that his version of the MSU tropospheric temperature trends, which are smaller than climate models predict, also agree with independent observations and therefore are more realistic than the Carl Mears version of the same data. Much of the difference appears to be related to the NOAA-9 (and to a lesser extent NOAA-11) target factor (merging parameters) in the intercallibration of the separate satellite instruments.
Tom Ackerman showed using raman lidar at the ARM Southern Great Plains site the existence of RH much higher than produced in climate models and occasions where no ice was detected even at RH=150%.
Junhong Wang talked about a new reference radiosonde system that can detect sub visible cirrus which may be important for understanding cloud feedback. The new sensor, SNOW WHITE, compared well with Vaisala and outperformed other older sondes such as the VIZ type. However, the ability of SNOW WHITE to accurately measure humidity in the upper troposphere and to detect thin cirrus cloud was argued to be a big advantage of the new sensors.
Remy Roca calculated Free Tropospheric Humidity (FTH - a better description than upper tropospheric humidity) using Meteosat 6.3 micron brightness temperatures and used different methods of cloud clearing in comparison with ECMWF data. He found the observations to be moister than ECMWF in the tropics and extra tropics (by about 3%), especially near to convective systems, although this may also be due to undetected cirrus in the obs. The obs were dryer than ECMWF in the subtropics by about 2%.
Tony Del Genio used precipitation and radiation data from the TRMM satellite along with NCEP 500mb vertical motion and HadISST SST and ISCCP D1 high cloud to constrain climate model parametrizations and assess the realism of the Ramanathan and Collins thermostat and Lindzen Iris hypotheses on tropical SST regulation. The analysis included sampling in vertical motion-SST space in a manner similar to that used by Mark Ringer (e.g. Williams et al., 2003). They found rain efficiency to increase with SST. The same occurs for ice precip efficiency up to 28 oC, above which efficiency falls again. Some non-convecting cloud increases with convective cloud (e.g. cirrus).
Bill Randel investigated the stratospheric water vapour and tropospheric temperature changes over the last 11 years using HALOE data and radiosondes. Drying at 83mb in the last 2 years are of the opposite sign of the previous moistening and are not understood.
B. J. Sohn presented a new method for calculating the combined cloud and water vapour forcing. This avoids the inconsistent clear-sky sampling in observations and models. See also Allan and Ringer (2003), submitted to GRL.
Brian Weare proposed a mechanism for the initiation of MJO by moisture variations: equatorially trapped propagating Kelvin Waves initiate convergence, building low level moisture through shallow convection which through increased CAPE initiates deep convection and upper tropospheric moistening. The moisture and latent heat anomalies lead OLR anomalies by ~10 days.
Brian Mapes argued that the observed bi-modal distribution of free tropospheric humidity could be explained by the mixing time-scale being longer than the subsidence drying time-scale.
Duane Waliser showed strong correlation between Monsoon variability and Intraseasonal Oscillation strength in models. The GFDL model often underestimates the INdian precipitation variability, in part due to an unrealistic double ITCZ here.
Dai and Trenberth showed a significant diurnal cycle of SST which is not included in AMIP models. Convective precip kicks off about 2 hours too early in the CCSM.
Gerry Potter presented a "poor-mans assimilation system" as a way of identifying simulation errors early. Using ARM/ERA40 RH, it was shown that the climate model used moistened the upper troposphere too readily.
Joel Norris used surface/ship observations of cloud to investigate and extend the recent evidence of large changes in the radiation budget from the 1980s-1990s (e.g. Wielicki et al. 2002). The ship data show decreases in upper cloud that are consistent with the increased OLR in the satellite data. By regressing the cloud and radiation budget data, changes in the tropical radiation budget since 1957 was inferred which show a 0.7% decrease in upper cloud, a 0.9-2.2 Wm-2 increase in OLR, a 0.5-1.2 Wm-2 decrease in RSW and a 0.4-1.0 Wm-2 decrease in net absorbed radiation over the 1957-1998 period.
Lazo used a Willingness to Pay method to determine the value of weather forecasts to the general public. $8.8 billion net benefit to the US public apparently.
Kevin Trenberth tried to reconcile an apprent seamless poleward energy transport with the sharp cut-off between the Hadley Circulation and Baroclinic Eddy transport at about 30oN/S using the top of atmosphere radiation budget. He argued that the SST distribution determines the position of the upward branch of the Hadley Circulation but not the descending node and that radiative ccoling in the descending branches do not drive the circulation contrary to the general view. The seamless energy transport is related to a compensation between the dry and moist static transport.
Antonio Navarra kicked the ECHAM4 model with x2, x4 and x16 CO2. The x16 CO2 run produced a permanent El Nino-like state.
Philander talked further on the possibility of permanent El Nino conditions in a warmer climate, arguing that it has happened before 40-60 Million years ago. The ENSO variability is possible only where the thermocline is shallow enogh for winds to initiate feedback cycles. 60 Myr ago the thermocline was too deep for this to occur. Temperature and salinity effects are required to model this well.
Alex Hall presented some interesting surface albedo feedback experiments. The albedo feedback was suppressed by prescribing seasonally varying surface albedo's into the radiative component of the GCM code. Present day variability and climate change experiments were undertaken for control (feedback) case and no albedo feedback cases. (NB The snow feedback is complicated because the absorbed surface radiation is most sensitive to the occurence of fresh snow for clear-sky conditions for highest insolation.) The surface albedo feedback amplifies the surface air temperature variability, paricularly over the S. Ocean and N.Hemisphere land. In Spring surface temperature is important for snow due to melting. At other times, the surface air temperature shows little dependence on the surface albedo feedback. Warming at high lats is much smaller without the albedo feedback, although there remains some polar amplification of the surface temperature response due to Sensible Heat feedback. However, the tropics is also sensitive to the albedo feedback.
My talk can be found here.
The best (only) decent pubs can be found at the corner of Pine Street and E Ocean Boulevard (Rock Bottom Brewery) and on the Pacific Terrace Harbour (The Yard House). In my opinion, probably the worst thing to do is order meatballs from Georges "Greek" restaurant (although they do serve Mythos which enables you to wash down the dire cuisine!).