Advancements in differential VLF: A low-cost approach to determining continuous lava effusion rates through a basaltic lava tube at Kilauea volcano, Hawaii using very low frequency electromagnetic monitoring

Continuous measurements of lava discharge, especially when output is hidden entirely within lava tubes, has proven extremely difficult. To overcome this problem, we have developed and tested a low-cost prototype instrument for continuously monitoring the cross-sectional area of lava in a master lava tube and estimating the instantaneous flux of lava flowing from a volcano, in this case, Kilauea volcano's East Rift Zone (ERZ), Hawaii. This design utilized two stationary very low frequency (VLF) radio receivers. One on the ground surface over a lava tube to measure the influence of highly conductive molten lava on a VLF signal transmitted from remote US military transmitters (ca. 400km distant). The second, some 50 m from the tube measures background VLF interference above solidified lava. The normalized difference in the VLF signals allows for the continuous monitoring of the cross-sectional area of molten lava in the lava tube and hence the name Differential VLF (DVLF) method. With velocity estimation, the instantaneous lava effusion rate can also be monitored. Data from a short, but continuous 4-hr test of the prototype DVLF instrument were compared against two discontinuous measurements taken by a hand-held Geonics EM-16, which initially measured the wet cross-sectional area of the tube as 11.7 m² and 65 minutes later at the time of the beginning of the DVLF measurements as 11.1 m². This 5% reduction is consistent with declining tilt observed on the ERZ at that time and demonstrates that the tube was only flowing at partial capacity. A plot of the difference in the amplitude of the DVLF signal received by our two VLF radios reveals evidence for variation in the cross-sectional area of lava flowing in the tube. A portion of this variation can be reasonably attributed to imperfect calibration, temperature drift and errors in the analog-to-digital process; however, these factors are in total very small and unlikely to produce the variations observed. Since it is unlikely for the slope and cross-section of the lava tube to vary significantly over the duration of the measurements, the data can be interpreted as proportionate to lava flux. Some of the variations observed have time scales of 5 to 60 minutes and are likely caused by fluctuating amounts of lava in the tube. This may represent changes in the lava effusion rate at a temporal scale not normally observed for lava tubes, but is consistent with observations of open channel flow described by others. A longer continuous time series of DVLF observation is needed. The next step is to compare the prototype DVLF instrument side by side with a more advanced but still low-cost DVLF instrument based on Stanford's A.W.E.S.O.M.E. receiver. The aim is to capture multiple cycles of known volcano inflation and deflation events over a 1 to 2 month duration, which accompany magma input to the shallow summit reservoir and eruption. While we cannot yet demonstrate the DVLF method's full capability at such short temporal scales conclusively, DVLF methodology represents a breakthrough in long-term, continuous monitoring capability of instantaneous effusion rate of an ongoing volcanic eruption at 10% to 20% the cost of the leading VLF instruments.