Implicit particle methods and their connection with variational 1 data assimilation

6 The implicit particle filter is a sequential Monte Carlo method for data assimilation that 7 guides the particles to the high-probability regions via a sequence of steps that includes 8 minimizations. We present a new and more general derivation of this approach and extend 9 the method to particle smoothing as well as to data assimilation for perfect models. We 10 show that the minimizations required by implicit particle methods are similar to those one 11 encounters in variational data assimilation, and we explore the connection of implicit particle 12 methods with variational data assimilation. In particular, we argue that existing variational 13 codes can be converted into implicit particle methods at a low additional cost, often yielding 14 better estimates that are also equipped with quantitative measures of the uncertainty. A 15 detailed example is presented. 16