We compare measurements for 1998 of VLF sferics-generated noise power at Halley, Antarctica, with the predictions of the long wave noise prediction (LNP) noise model. We compare the highest-ranked predicted sources of radio noise for Halley from LNP with the observed lightning flash rates in January and July 1998 observed by the Lightning Imaging Sensor (LIS) instrument on the Tropical Rainfall Measurement Mission (TRMM). In both cases we find better agreement between LNP and the measurements in southern winter than those in southern summer. We compare an anomaly time series derived from VLF power with an analogous temperature anomaly series. We find evidence of a change in measured power in 1998 compared to previous years, suggesting that the tropical temperature signal of the strong El Niño of 1997–1998 may be detectable in VLF power measurements at Halley. Assuming that such an event and a single temperature station can be used to calibrate VLF measurements as a “tropical thermometer”, we confirm previous estimates that long-term changes in tropical noise measured at Halley are consistent with a tropical temperature rise of the order of ∼ 1°C over 25 years.