HR: 0800h
AN: GC51A-0154    [Abstracts]
TI: Tropical Lapse Rates From Radiosondes: Does a Model-Obs Discrepancy Exist?
AU: * Titchner, H A
EM: holly.titchner@metoffice.gov.uk
AF: Met Office Hadley Centre, FitzRoy Road, Exeter, EX1 3PB, United Kingdom
AU: Thorne, P W
EM: peter.thorne@metoffice.gov.uk
AF: Met Office Hadley Centre, FitzRoy Road, Exeter, EX1 3PB, United Kingdom
AU: McCarthy, M P
EM: mark.mccarthy@metoffice.gov.uk
AF: Met Office Hadley Centre, FitzRoy Road, Exeter, EX1 3PB, United Kingdom
AB: Debate continues over tropical tropospheric temperature trends during the satellite era. In particular, over whether reported trend behaviour supports climate models which robustly show the tropical troposphere acting as a pseudo-moist adiabat. To date the structural uncertainty and potential systematic bias that results from the different choices that could be made during dataset construction are poorly quantified. We have developed an automated dataset homogenisation system for assessing these issues in HadAT, the Met Office Hadley Centre radiosonde temperature dataset (available at www.hadobs.org). Our system is able to produce an ensemble of dataset realisations. Here we assess our ability to recover the true long term tropical trends by analysing simulated data from an atmosphere only GCM run with prescribed SSTs and human and natural forcings with four very different sets of complex biases super-imposed. As we know the answer in each case these test case ensembles can be used to benchmark the system capabilities and then reinterpret the tropical trends that the system outputs when it is given the raw observations as input (where the truth is unknown). In this work we produce ensembles containing 100 members. Our results strongly imply that most current radiosonde and some satellite datasets exhibit too little warming aloft within the tropics. Although we can place a strong lower bound on the satellite era trends (0.05K/decade) this lower bound still allows for divergent behaviour between models (0.14-0.20K/decade) and the real world. That being said, we cannot place a defensible upper bound on trend estimates and therefore our uncertainty range does include this model expectation.
DE: 1616 Climate variability (1635, 3305, 3309, 4215, 4513)
DE: 3305 Climate change and variability (1616, 1635, 3309, 4215, 4513)
DE: 3374 Tropical meteorology
SC: Global Environmental Change [GC]
MN: 2007 Fall Meeting