Improvement of Tidal Analysis Results by a Priori Rain Fall Modelling at the Vienna and Membach stations

We investigate how far tidal analysis results can be improved when a rain fall admittance model is applied on the superconducting gravity (SG) data. For that purpose both Vienna and Membach data have been analysed with and without a priori rain fall correction. In Membach the residual drop for most events (80%) can be explained by the rain water load, while in Vienna only 50% of all events fit the model in detail. In the other cases the Newtonian effect of vertical air mass redistribution (vertical density variation without air pressure change), predominantly connected with high vertical convection activity, e.g. thunderstorms, plays an essential role: short-term atmospheric signals show up steep gravity residual decreases of a few nms-2 within 10 - 60 min, well correlated with outdoor air temperature in most cases. However, even in those cases the water load model is able to explain the dominating part of the residual drop especially during heavy rain fall. In Vienna more than 110 events have been detected over 10 years. 84% of them are associated with heavy rain starting at or up to 10 min later than the residual drop while the rest (16%) shows no or only little rainfall. The magnitude of the gravity drop depends on the total amount of rainfall accumulated during the meteorological event. Step like signals deteriorate the frequency spectrum estimates. This even holds for tidal analysis. As the drops are of physical origin, they should not be eliminated blindly but corrected using water load modeling constrained by high temporal resolution (1 min) rain data. 3D modeling of the water mass load due to a rain event is based on the following assumptions: (1) Rain water intrudes into the uppermost soil layer (close to the topography surface) and remains there at least until rain has stopped. This is justified for a period of some hours after the rainfall as evapotranspiration is not yet effective. (2) No run-off except of sealed areas or building roofs, where water can not intrude into the soil but will drain off into the sewage water system instead. (3) Rainfall is equal everywhere in the station surroundings. (4) No surface deformation due to the water mass load Correcting for rain fall effects reduces by about 10% the standard deviation of the residuals after tidal parameter adjustment. Amplitude factor changes are in the order of 10-3 or less, phase lags change by 10-3 to 10-2: statistically, these variations are not significant as they lie within the error bars. However, it is worth noting that the amplitude factors of tidal constituents with high amplitude (O1, P1, K1) and even Ψ1 and Φ1 show similar variations in Vienna and Membach. Generally the tidal parameter variation is less in the SD than in the D band.