Uncertainty remains as to what extent variability in mid to upper tropospheric moisture, especially over the tropics, behaves as constant relative humidity during interannual climate variations associated such as ENSO. Systematic variations in relative humidity suggest from HIRS 6.7 micron and MLS 205 GHz suggest that dry subtropical regions evolving during warm SST events depress relative humidity, but the dynamical interpretation of these events is still uncertain. Additional specific concerns have to do with regional signatures of convective processes: How does the origin of dry air in the eastern subtropical N. Pacific differ in ENSO warm versus cold years? The dynamics of Rossby wave forcing by convective heating, subtropical jet stream dynamics, and radiatively driven subsidence all come into play here. How variations in precipitating ice hydrometeors from tropical anvils relate to variations in UTH is also a subject of debate. Do variations in precipitating ice, cloud cover and water vapor behavior show any support for the Iris-hypothesis mechanism? Here we examine historical records of SSM/T-2 data to gain a better physical understanding of the effects of deep convective moisture sources and dynamically-induced vertical circulations on UTH. These high frequency microwave measurements (183.3 GHz) exhibit far less sensitivity to cloud hydrometeors than the 6.7 micron data and enable a potentially more robust record of upper tropospheric relative humidity. Furthermore, signatures of precipitating ice from these channels facilitate comparisons to TRMM hydrometeors detected by radar. In analyzing these observations, we isolate water vapor and temperature change components that affect 183.3 brightness temperatures and the inferred relative humidity. Trajectory modeling is also used to understand interannual humidity anomalies in terms of outflow from convective regions and history of diabatically-driven sinking which modifies relative humidity.