https://archive.ipcc.ch/publications_and_data/ar4/wg1/en/ch3s3-4-2-2.htmlIn general, the radiosonde trends are highly suspect owing to the poor quality of, and changes over time in, the humidity sensors (e.g., Wang et al., 2002a). Comparisons of water vapour sensors during recent intensive field campaigns have produced a renewed appreciation of random and systematic errors in radiosonde measurements of upper-tropospheric water vapour and of the difficulty in developing accurate corrections for these measurements (Guichard et al., 2000; Revercombe et al., 2003; Turner et al., 2003; Wang et al., 2003; Miloshevich et al., 2004; Soden et al., 2004).
https://www.ipcc.ch/site/assets/uploads/2018/07/WGI_AR5.Chap_.2_SM-1.pdf2.SM.6.2 Radiosonde Humidity Data
Since AR4 there have been three distinct efforts to homogenize the
tropospheric humidity records from operational radiosonde measurements (Durre et al., 2009; McCarthy et al., 2009; Dai et al., 2011) (Table
2.SM.9)
Abstract
[1] In an effort to update previous analyses of long‐term changes in column‐integrated water vapor, we have analyzed trends in surface‐to‐500‐hPa precipitable water (PW) calculated from radiosonde measurements of dew point depression, temperature, and pressure at approximately 300 stations in the Northern Hemisphere for the period 1973–2006. Inhomogeneities were addressed by applying a homogenization algorithm that adjusts for both documented and undocumented change points. The trends of the adjusted PW time series are predominantly upward, with a statistically significant trend of 0.45 mm decade−1 for the Northern Hemisphere land areas included in the analysis. Particularly significant increases are found in all seasons over the islands of the western tropical Pacific, and trends are also positive and statistically significant for the year as a whole and in at least one season in Japan and the United States. These results indicate that the widespread increases in tropospheric water vapor, which earlier studies had reported and shown to be physically consistent with concurrent increases in temperature and changes in moisture transport, have continued in recent years.
Radiosonde‐based trends in precipitable water over the Northern Hemisphere: An update
Imke Durre Claude N. Williams Jr. Xungang Yin Russell S. Vose
First published:13 March 2009
Abstract
A new analysis of historical radiosonde humidity observations is described. An assessment of both known and unknown instrument and observing practice changes has been conducted to assess their impact on bias and uncertainty in long-term trends. The processing of the data includes interpolation of data to address known sampling bias from missing dry day and cold temperature events, a first-guess adjustment for known radiosonde model changes, and a more sophisticated ensemble of estimates based on 100 neighbor-based homogenizations. At each stage the impact and uncertainty of the process has been quantified. The adjustments remove an apparent drying over Europe and parts of Asia and introduce greater consistency between temperature and specific humidity trends from day and night observations. Interannual variability and trends at the surface are shown to be in good agreement with independent in situ datasets, although some steplike discrepancies are apparent between the time series of relative humidity at the surface. Adjusted trends, accounting for documented and undocumented break points and their uncertainty, across the extratropical Northern Hemisphere lower and midtroposphere show warming of 0.1-0.4 K decade(-1) and moistening on the order of 1%-5% decade (1)since 1970. There is little or no change in the observed relative humidity in the same period, consistent with climate model expectation of a positive water vapor feedback in the extratropics with near-constant relative humidity.
An Analysis of Tropospheric Humidity Trends from Radiosondes
Article (PDF Available) in Journal of Climate 22(22):5820-5838 · November 2009 with 76 Reads
DOI: 10.1175/2009JCLI2879.1
Cite this publication
Mark P. McCarthy Et.al
One could look towards data we believe is reasonably accurate rather than relying on that we know is not.