Representation and diurnal variation of upper tropospheric humidity in observations and models
Abstract: The role of water vapour is manifold in its climate regulation of the Earth system. Most important of all despite its low concentration, is the role it plays in the upper troposphere. It assumes an important role in its contribution to greenhouse warming by way of its positive feedback effect, amplifying the radiative forcing due to increasing CO2 concentrations. Understanding the variability and distribution is thus important from a climate point of view and critical because the challenges involved in it are far too many. This thesis consists of an introduction and three research articles focusing on the study of upper tropospheric humidity (UTH). The first two articles are on two important sources of UTH data, the radiosondes and satellite data, and the third is associated with climate models, important tools for simulating and reproducing global climate features. The summaries of these three articles are as follows: Radiosondes have been the primary sources for vertical profiles of various atmospheric parameters and are one of the crucial components in numerical weather predictions and satellite validations. However, they are known to have certain issues withmeasurements of humidity in the upper troposphere. The first article highlights the importance of radiosonde humidity corrections by using satellite measurements as the reference. The infrared and microwave measurements from NOAA-17 polar orbiting satellite were used as the reference in this study. Collocated radiosonde measurements from the Atmospheric Radiation Measurement (ARM) campaign were converted into satellite radiances using the ARTS radiative transfer model. The comparisons with satellite measurements were done separately for daytime and nighttime soundings of radiosonde under clear sky conditions. An empirical correction procedure meant to address the mean bias error and solar radiation error was applied to the radiosondes. The empirical correction was found to significantly reduce the dry bias of radiosondes in the upper troposphere. The impact of the correction is prominent over daytime radiosonde measurements on account of the bias removal associated with the solar radiation error. Long term time-series measurements of tropospheric humidity are available from polar orbiting satellites but the measurements from these satellites have been found to be affected by diurnal sampling bias, which is caused by a drift in the orbital height of the satellites, thus changing the local sampling time of the satellites over course of time. This therefore introduces a spurious trend into the time-series data obtained from these satellites. A methodology for the correction of orbital drift error applied on microwave humidity measurements from NOAA and MetOp-A satellites forms the subject of the second article included in this thesis. Climatological diurnal cycles of microwave humidity measurements were obtained from 5 different polar orbiting satellites to infer and thereby correct the diurnal sampling bias in microwave humidity measurements. The diurnal cycles were generated separately for the 5 microwave channels. A Monte Carlo error analysis also determines the significance of diurnal amplitudes. The impact of diurnal correction has been evaluated by analyzing the surface channel brightness temperature time-series of NOAA-16 and UTH channel time-series of NOAA-17 satellites. The impact of diurnal correction is greater for the surface channels when compared to the UTH channels due to the larger diurnal cycle amplitudes in the surface channels. Climate models are one of the main tools for the prediction of future climate change. Most processes associated with water vapour appear in climate models as parameterizations since they are too small-scale or complex to be physically represented in models. Therefore, frequent validation of models against observations is needed to assure their reliability. The third article evaluates the performance of two climate models, in simulating the diurnal cycles of upper tropospheric humidity taking combined microwave humidity measurements from four different satellites as the reference. The comparisons were made over the convective land and oceanic regions over the tropics. The diurnal cycle differences between infrared and microwave observations and the reason for these differences are also analyzed. It is shown that the cloud sensitivity differences in infrared data can shift the diurnal phase relative to microwave data. The models exhibit considerable differences in the diurnal phase and amplitude of UTH as against microwave observations over both land and oceanic regions.
This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.