FREQUENCY RESPONSE OF THE STREAM AND SOIL TO RAIN INPUT AT MINK BROOK, ETNA, NH

We have been collecting high temporal resolution data for precipitation, soil moisture (down to 50cm), and streamflow at Mink Brook, Etna, NH, since the fall of 2002 in order to investigate mechanisms of hydrological response to precipitation events. The resolution of the data allows for the application of spectral analysis. Kirchner et al. (2000) used spectral analysis to show that the concentration of "conservative" species in rain is damped by the watershed into inverse frequency noise. Streamflow was not significantly damped for frequencies less than 1/day. Our higher resolution data have shown that frequencies greater than 1/day are damped in the soil moisture and streamflow record. High frequency rain fluctuations up to 15/day are transmitted to the stream but do not show up in the soil moisture records deeper than 20 cm. One explanation could be that the soil moisture amount does not show a response in slight pressure changes which are transmitted through the unsaturated zone. Alternatively, preferential pathways could be bypassing the soil probes.

We compare two time periods, both having similar soil moisture contents in the upper unsaturated zone but differing in the stream baseflow. The period with the lower baseflow shows higher frequency signals transmitted to the stream but no difference in the soil response. This response to higher frequency input under dry conditions may be due to only the shortest flowpaths connecting to the stream when baseflow is low. One significant pathway is direct precipitation on the stream (Renshaw et al., 2003). Under wetter conditions, activation of additional transport pathways will result in a wider distribution of travel times to the stream and a further loss of high frequency temporal variations in stream discharge. The activation of longer flow paths will also push out “old” water that has been in the watershed, damping the chemical signal of “new” input water. These long flowpaths can be activated during light rains when baseflow is high or alternatively during times of heavy rain. This leads to a somewhat counterintuitive conclusion that heavier rains will likely result in more dilution of the “new” water signal in the stream as has been observed by Renshaw et al. (2003).