The premise of this work is the development and application of a new methodology to forecast production data in unconventional reservoirs where variable rate and pressure drop data are typically observed throughout production. Decline curve analysis techniques for the estimation of ultimate recovery (EUR) require the constant bottomhole pressure condition during the producing life of the well-whereas it is not regular practice to maintain a constant bottomhole pressure profile throughout production in unconventional reservoirs. Therefore, the applicability of the time-rate decline relations is questionable, and methods to remove pressure variations from rate response are needed for generating future production forecasts. From a conceptual view point, we propose the utilization of the convolution/superposition theory along with the recently developed "empirical" time-rate equations, which are normalized by pressure drop data. In order to avoid non-uniqueness, a workflow is used where model parameters for the "normalized" decline curve equations are identified using diagnostic "qDb" plots. Normalized decline curve equations are then convolved with the pressure drop data to achieve a history match and to forecast production. We provide demonstrative application of this technique using an example from an high pressure high temperature shale gas reservoir. For varying bottomhole pressure cases, we show that our proposed techniques effectively remove pressure variations from the rate history. We present the differences in computed EUR values using decline curve analysis with and without corrections for varying pressures. In addition, forecasts are generated using supplementary plots such as pressure drop normalized rate versus cumulative production.