Multi-Technique Analysis of Precipitable Water Vapor Estimates in the sub-Sahel West Africa

TL;DR

This study evaluates precipitable water vapor in sub-Sahel West Africa using AERONET data, calibrates empirical formulas, and assesses the accuracy of NWP models, revealing model overestimations and proposing a region-specific formula.

Contribution

It introduces a calibrated empirical formula for PWV estimation specific to sub-Sahel West Africa and evaluates NWP model performance in this region.

Findings

Empirical formula PWV=aT_d+b fits observed data well.

NWP models overestimate PWV by about 25%.

AERONET data effectively calibrates humidity-based PWV estimates.

Abstract

Precipitable water vapor (PWV) is an important climate parameter indicative of available moisture in the atmosphere, it is also an important greenhouse gas. Observations of precipitable water vapor in sub-Sahel West Africa are almost non-existent. Several Aerosol Robotic Network (AERONET) sites have been established across West Africa, and observations from four of them, namely, Ilorin (4.34o E, 8.32o N), Cinzana (5.93o W, 13.28o N), Banizoumbou (2.67o E, 13.54o N) and Dakar (16.96o W, 14.39o N) are being used in this study. Data spanning the period from 2004 to 2014 have been selected, they include conventional humidity parameters, remotely sensed aerosol and precipitable water information and numerical model outputs. Since in Africa, only conventional information on humidity parameters is available, it is important to utilize the unique observations from the AERONET network to calibrate empirical formulas frequently used to estimate precipitable water vapor from humidity measurements. An empirical formula of the form PWV=aT_d+b where T_d is the surface dew point temperature, a and b are constants, was fitted to the data and is proposed as applicable to the climatic condition of the sub-Sahel. Moreover, we have also used the AERONET information to evaluate the capabilities of well-established numerical weather prediction (NWP) models such as ERA Interim Reanalysis, NCEP-DOE Reanalysis II and NCEP-CFSR, to estimate precipitable water vapor in the sub-Sahel West Africa, it was found that the models tend to overestimate the amount of precipitable water at the selected sites by about 25 %.