Water vapor measurements in the Arctic middle atmosphere by means of a novel ground-based microwave spectrometer

Abstract

This Ph.D. thesis is centered on the development and improvement of the acquisition software and the data analysis algorithms of the ground-based 22.23 GHz spectrometer VESPA-22 (water Vapor Emission Spectrometer for Polar Atmosphere at 22 GHz), on the instrument installation and on the analysis of the first year of data collected. The instrument measures the 22.23 GHz water vapor emission line with a bandwidth of 500 MHz and a frequency resolution of 31 kHz, in order to retrieve the middle atmospheric water vapor profile and the atmospheric opacity. The long time series of measurements provided by a ground spectrometer is a useful tool to study the Arctic climate, the atmospheric chemical composition and the air mass dynamic. The tropospheric water vapor is the major responsible for the greenhouse effect on the planet and also its stratospheric component has a measurable effect on the atmosphere and the surface radiation budget (Solomon et al., 2010). Water vapor is also related to many chemical processes occurring in the Arctic stratosphere, such as the ozone depletion (Solomon, 1999). The stratospheric water vapor, due to its long lifetime, can be also used as a tracer to study dynamic processes such as the polar vortex (Lahoz et al., 1993). The long term series of measurements of water vapor vertical profiles by ground-based spectrometers are particularly valuable to derive trends (Nedoluha et al., 1999). VESPA-22 is at the moment installed at the THAAO (Thule High Arctic Atmospheric Observatory) located at Thule Air Base, Greenland (76.53°N, 68.70°E). It is designed to operate automatically with minimum need of maintenance; it employs an uncooled front-end characterized by a receiver temperature of about 180 K and its quasi-optical system presents a full width at half maximum angle of 3.5°. The instrument calibration is performed automatically by a noise diode; the emission temperature of this element is measured two times an hour through the observation of a black body at ambient temperature and of the sky at 60° of elevation. The retrieved profiles obtained inverting a 24-hour integration spectra present a sensitivity higher than 0.8 from about 25 to 72 km of altitude, a vertical resolution from about 12 to 23 km (depending on altitude) and an overall 1 uncertainty between 5 and 12 %. During the Ph.D. work, I have expanded, automated and further developed the VESPA-22 acquisition software, allowing the instrument to perform all the routine tasks in a remote location without needing of maintenance by local personnel. The instrument and the acquisition software proved to work reliably: VESPA-22 collected measurements for more than a year with few interruption periods characterized by very poor weather. I also improved the measurement procedures used by the instrument and data analysis codes. In particular, many efforts of this work were directed to the development of a reliable retrieval algorithm, used to obtain the vertical water vapor concentration profiles from the measured spectra. I participated in the VESPA-22 installation campaign occurred in July 2016 at Thule and to a preparatory campaign which took place on February-March 2016 in order to setup the THAAO laboratory to host the instrument. I analyzed the first year of data of the instrument, from July 2016 to July 2017 and compared VESPA-22 and satellite instrument Aura/MLS (Waters et al., 2006) water vapor datasets in order to evaluate VESPA-22 results. In the sensitivity range of VESPA-22 retrievals, the intercomparison between VESPA-22 dataset and Aura/MLS dataset convolved with VESPA-22 averaging kernels reveals a correlation coefficient of about 0.9 or higher and an average difference reaching its maximum of -5% or -0.2 ppmv at the top of the sensitivity range. I employed the VESPA-22 opacity measurements collected during fair weather to obtain an estimate of precipitable water vapor (PWV) through a fit with the measurements of the HATPRO (Humidity And Temperature Profilers) radiometer (Rose and Czekala, 2007) installed at the THAAO, as described by the work of Deuber et al. (2005). The difference between the PWV values measured by HATPRO and estimate by VESPA 22 allows the estimation of the uncertainty of this technique, equal to 8%