A fixed‐layer estimation of upper tropospheric specific humidity from the GOES water vapor channel: Parameterization and validation of the altered brightness temperature product

Abstract

A derived‐product image (DPI) of specific humidity is determined at the pixel‐level using the GOES 6.7 μm water vapor channel and ancillary model temperature fields. By treating the brightness temperature, T6.7, as a composition of three effectively independent influences (specific humidity, temperature and zenith angle), the ancillary temperature fields allow an estimation of the sole effect of specific humidity on the variance of T6.7, primarily in the midlatitudes. This variance is approximately linear with the logarithm of a vertically weighted average of specific humidity. As an intermediate step, we have created a term called the altered brightness temperature product, Talt6.7, that represents the variance of T6.7 due to specific humidity plus an approximate mean of T6.7 in the midlatitudes. The domain of this derived product extends to zenith angles often greater than 70° within the GOES field of view. However, strong vertical gradients in specific humidity in the midtroposphere to upper troposphere were found to violate the linearity of the regression. In the validation, the slope of the first‐order regression is robust but the constant is inconclusive, rendering the product better for relative rather than absolute associations of specific humidity. Comparison to potential vorticity (PV) shows that the derived product corresponds best to PV at the 300 hPa pressure level at values of PV associated with the tropopause. The combined use of the DPI and model‐derived PV contours yields an effective tool for analyzing stratosphere‐troposphere exchange in the midlatitudes.