## Permeability for Columbia River Basalt Group, USA

Hydraulic conductivity estimatesfor CRBG aquifers were taken from the USGS National Water Information System (NWIS) database (Kahle et al. 2011) and converted to permeability estimates using temperature estimates from the nearest temperature profile log (Fig. 2). Temperature was estimated by using a best linear fit to the 100 m of borehole temperature log closest to the depth of the aquifer test. Viscosity and density were estimated as a function of temperature, and hydraulic conductivity was converted to permeability by multiplying by viscosity and dividing by the product of density and the gravitational acceleration constant (Fetter 1994, p. 96). Bulk permeability was computed using the layered system approximation (the arithmetic mean), assuming that the aquifer occupies 10% of the total thickness and that the permeability of the lava flow interiors is negligibly small.

The detailed CRBG aquifer transmissivity estimates of Spane (2013) were also converted to estimated permeability (Fig. 4). First, estimated transmissivity was converted to bulk hydraulic conductivity by dividing by 30 m or the open length of the borehole, whichever was longer. Short open boreholes were assumed to test individual aquifers, so a typical lava flow thickness of 30 m was assumed to convert the aquifer test to a bulk permeability by using the layered system approximation (Fetter 1994, p. 123–124). Long open borehole tests (>30 m) potentially intersected multiple aquifers and a representative length of confining unit. Temperature measured in these boreholes (Schroder & Strait 1987; McGrail et al. 2009; Spane et al. 2012) was used to correct viscosity and density, and permeability was computed and assigned to the middle of the test interval. Four boreholes have multiple packer tests vertically along the borehole length.