The 19 March 2017 tsunami at the Port of Dayyer incurred considerable damage and resulted in ~$10 million of economic loss, one death and five people missing. Although a significant storm had preceded the surge, there were indications of calm weather during the event in some of the remaining footage which could play a critical role in reconstructing the tsunami by studying it as a meteorological phenomenon. Here, we are reporting the data from a field survey in the affected area which revealed a concentration of inundation along a ~30 km stretch of coastline with run-ups reaching 3 m at ~4 km west of Dayyer. In the absence of any major seismic event at or around the occurrence of the tsunami, we consider landslides as well as atmospheric phenomena as potential sources for the event. We simulate the geological tsunami using the Method of Splitting Tsunamis (MOST) (e.g. Titov et al., 2016) and develop a finite-difference algorithm, following the method by Platzman (1958), to model the surge as a meteotsunami. Our simulations suggest that the Dayyer tsunami was caused by a local system of atmospheric pressure gradient, as our landslide models fail to reproduce high amplitudes as documented in our field survey.