Quantitative Precipitation Forecasting (QPF) forms an important aspect of atmospheric as well as hydrologic prediction, and has been a topic of extensive research over the last few decades. Comparing model output to observations (QPF verification) is not trivial, owing to (i) the considerable variability that precipitation exhibits over a wide range of scales; and (ii) the discrepancies in resolution between various sources of precipitation estimates and model output. Recent work in our group has documented evidence suggesting that the typical procedures of interpolating model output to the scale of observations or vice-versa are deficient in that they impose a ``representativeness error'' that is significantly different from zero, even when the model is considered to be perfect. To alleviate this concern, the use of a procedure called Scale-Recursive Estimation (SRE), based on Kalman filtering, was proposed. Given precipitation observations and the associated uncertainties at different scales (e.g., from raingauges, radars and satellites), the method allows for explicitly incorporating a known multiscale structure (analytical or empirical) of precipitation such that the best (unbiased) estimate of a field at any desired scale (for instance, at the scale of a model output), along with its uncertainty, can be obtained. While our previous work dealt with showing the utility of such a method using a proof-of-concept approach, the current work presents a case-study of the verification of precipitation obtained from a numerical high-resolution model, the Advanced Regional Prediction System (ARPS).