Central U.S. Atmospheric Water and Energy Budgets Adjusted for Diurnal Sampling Biases Using Top-of-Atmosphere Radiation

Abstract

The water and energy budgets of the atmospheric column over the Mississippi River basin are estimated using 18 yr (1976?93) of twice-daily radiosonde observations, top-of-atmosphere net radiation estimates from the Earth Radiation Budget Experiment (ERBE), streamflow measurements, and reanalysis-based estimates of ground heat storage. The seasonal cycle of the divergence of moisture, moist static energy, and dry static energy are decomposed into three terms, the mean divergence, mean advection, and eddy source, each of which are seen to have similar magnitudes and, in many cases, opposite signs. Because of this, the residuals found when comparing atmospheric energy divergence to top-of-atmosphere net radiation or when comparing moisture divergence with land surface streamflow, are highly sensitive to propagating errors. Further analysis of four times per day radiosonde observations from an earlier 18-yr period (1948?65) indicate that the divergence profile of the wind field oscillates equally strongly at seasonal, diurnal, and semidiurnal time scales. The implication of these variations is that climatological means formed by twice-daily sampling (0000 and 1200 UTC) will produce a systematic, seasonally varying bias in the water and energy budgets. To estimate this bias, the monthly mean wind divergence profile is adjusted from a prior estimate until the atmospheric energy budget is closed. The divergence of the eddy and mean-advection terms, which showed significantly less diurnal variability, are unaltered. As an independent check on the procedure, the moisture budget calculated using the adjusted wind divergence profiles are compared with streamflow-based estimates of terrestrial runoff. The adjustment procedure increases the moisture convergence from 0.2 to 0.5 mm day?1, bringing it closer to the terrestrial estimate (0.6 mm day?1), and induces a qualitatively more reasonable looking seasonal cycle in the mean-divergence term.