WELLVEL: WELLington (New Zealand) region water WELLbore levels and seismic VELocity changes

Seismic wave speeds, water bore levels and GNSS station locations all changed at the time of the 2016 Kaikoura earthquake. The WELLVEL project is determining the relation between these types of changes by re-examining data from permanent GeoNet stations and by deploying temporary seismometers at sites of 11 water bores in the Wellington region that were evaluated for changes due to the Kaikoura earthquake. Three started in March 2020, just before the Covid-19 lockdown, and the rest were installed in July through September, 2020. For the permanent data on the South Island, using the station-to-station noise cross-correlation technique we have found drops in velocity at the time of the Kaikoura earthquake ranging from 0-0.2% for paths between stations that had a natural period of 2 Hz, and an average of 0.15% for paths between broadband stations. The velocities at the broadband stations also dropped about 0.1% over the time of the Cook Strait earthquakes. Using cross-component correlations on stations in the Wellington region, we measure velocity drops up to 0.5% on southwestern stations BHW and WEL at the time of the Kaikoura earthquake for the frequency range 0.1-1 Hz. These drops are not as strong on stations located further north, with the largest amplitude drop of about 0.25% on station KIW at Kapiti Island and OGWZ just northeast of KIW. The 5th March 2021 M7.3 Te Araroa earthquake about 600 km northeast of Wellington produced the most intense ground shaking response on the temporary network, with an initial 174 mm drop in water height and a peak ground acceleration of 3.9 cm/s^2. This event and two more distant Kermadec earthquakes (M=8.1 and 7.4) on the same day produced tsunamis that were visible on a wellbore near the shore. However, none of the stations showed any net increase or decrease in water level or in seismic velocity in the hours following the event. For longer-term environmental signals over several months, we observed direct correlation between water height and seismic velocity at some sites, but inverse correlations or no correlation at others. We are checking our processing parameters to see if we can get stronger signals to be sure whether we can distinguish the difference between signal and noise when we compare velocity changes to earthquake occurrence and to water well levels.