Using Limited Time Periods as a Means to Elucidate Microwave Sounding Unit Derived Tropospheric Temperature Trend Methods

摘要

Limited Time Period (LTP) running trends are used to elucidated Microwave Sounding Unit (MSU) derived tropospheric temperature trend methods. Regression derived coefficients were used to combine lower stratosphere (LS) and mid-troposphere to lower stratosphere (MT) simulated MSU channels from RATPAC radiosonde data using sites not found to have long-term cooling biases. This combination is used to estimate tropospheric temperature trends and then is compared to the actual RATPAC derived tropospheric temperature trends. It is found that statistical LS/MT combination results in greater than 30% error over some LTP. These errors are found to exist when strong cooling in the stratosphere is coincident with the zero trend level above the tropopause. LTP trends are also created from various MSU difference time series between the University of Alabama in Huntsville (UAH) and Remote Sensing System (RSS) group's lower troposphere (LT) and MT channels. This is accomplished in an effort to determine the causes of the greatest discrepancies between the two data sets. Results indicate the greatest discrepancies were over time periods where NOAA- 11 through NOAA-15 adjustments were applied to the raw LT data over land. Discrepancies in the LT channel are shown to be dominated by differences in diurnal correction methods due to orbital drift; however, discrepancies from target parameter differences are also present. Comparison of MSU data with radiosonde data indicate that RSS's method (use of climate model) of determining diurnal effects is overestimating the correction in the LT channel. Diurnal correction signatures still exist in the RSS LT time series and are likely affecting the long term trend with a warm bias. Using these findings it is shown that atmospheric amplification is not happening in globally averaged observations over the MSU era.