The Relationship Between Surface and the Lower Troposphere Anomalies

The observation that surface and atmospheric temperature trends are different over the past 20 years has sparked a lot of controversy and uncertainty. This discrepancy has led to numerous questions about the accuracy and precision of the various data sets used to monitor these parameters. Our study will investigate the possibility that the disparity between the surface and atmospheric records may be partially related to these parameters responding differently to the current radiative flux between the earth's surface and the lower troposphere. Although it is highly unlikely that such a divergence could continue for an extendedperiod, it is possible that short term (decadal time scale) anomalously warm or cold periods in various regions of the world can be translated differently between the atmosphere and surface. This is particularly true when boundary layer inversions are pervasive or differential patterns of latent heat release alter the pattern of the energy transfer. Our study will investigate the reality of this decoupling and its correspondence to changes in the global temperature patterns. We are able to perform this analysis with improved spatial resolution through the development of a new data set that blends land surface temperature anomalies from in situ and satellite observations into one superior product. This land surface product is then merged with NOAA’s blended sea surface temperature product. The final data set provides near global coverage of surface temperature anomalies at one degree resolution. We use this data set in conjunction with MSU and CARDS tropospheric temperature anomalies to analyze the surface and atmospheric correspondence and decoupling. The blended surface data is currently available for the period 1992- 1999, allowing a long enough record to identify periods and areas where this relationship changes. We will also determine how this variability is associated with the phases of the southern oscillation. The short period of record does limit its utility for determining how these relationships change over an extended time period. We will propose a scheme to overcome this limitation, with hopes the audience can provide useful feedback on our approach.