Deriving meteor stream properties from meteor count rate time series in a multi-observer network

Meteor count rate time series collected during showers are a prime source of data for characterizing meteor streams. Doing so requires corrections for the sporadic meteor background, for the radiant positions in the sky, and for the specific properties of the observing equipment. This paper presents a suite of methods for analyzing meteor shower count rate time series using variants of the single- and double-exponential shower activity model. The approach relies on a least-squares fitting procedure. It is demonstrated that the basic version of the problem is often ill-posed. Constraining the fitting procedure by making assumptions about the sporadic background offers a partial remedy. To allow for a more robust solution, however, a generalization is developed that includes data from multiple observers. This makes the technique especially powerful when analyzing the data from forward scatter radio meteor networks. A Monte Carlo approach allows to establish confidence intervals on the results obtained. A selection of results illustrates the capabilities and limitations of these methods. • We have elaborated a methodology to disentangle sporadic and stream meteor contributions from time series of meteor counts, and to recover the characteristics of the corresponding meteor stream(s), including uncertainty margins on them. • The methodology supports an array of models for the sporadic meteor activity, for the observability functions, and for the stream meteor intensity, to maximize the ratio of the number of observations over the number of unknowns, while improving the realism of the solution. • Applying these methods to forward scatter radio meteor data for the 2019 Geminids, we recover the typical asymmetry in activity rise and decay time scales.