A major set of changes was made to the Environment Canada global deterministic prediction system during the fall of 2014, including the replacement of four-dimensional variational data assimilation (4DVar) by four-dimensional ensemble–variational data assimilation (4DEnVar). The new system provides improved forecast accuracy relative to the previous system, based on results from two sets of two-month data assimilation and forecast experiments. The improvements are largest at shorter lead times, but significant improvements are maintained in the 120-h forecasts for most regions and vertical levels. The improvements result from the combined impact of numerous changes, in addition to the use of 4DEnVar. These include an improved treatment of radiosonde and aircraft observations, an improved radiance bias correction procedure, the assimilation of ground-based GPS data, a doubling of the number of assimilated channels from hyperspectral infrared sounders, and an improved approach for initializing model forecasts. Because of the replacement of 4DVar with 4DEnVar, the new system is also more computationally efficient and easier to parallelize, facilitating a doubling of the analysis increment horizontal resolution. Replacement of a full-field digital filter with the 4D incremental analysis update approach, and the recycling of several key variables that are not directly analyzed significantly reduced the model spinup during both the data assimilation cycle and in medium-range forecasts.

The spectral-difference algorithm of Barker and Marshak for inferring optical depth of broken clouds has been shown numerically to be potentially useful. Their method estimates cloud-base reflectance and using spectral radiometric measurements made at the surface at two judiciously chosen wavelengths. Here it is subject to sensitivity tests that address the impacts of two ubiquitous sources of potential error: instrument noise and presence of aerosol. Experiments are conducted using a Monte Carlo photon transport model, cloud-resolving model data, and surface albedo data from satellite observations. The objective is to analyze the consistency between inherent and retrieved values of . Increasing instrument noise, especially if uncorrelated at both wavelengths, decreases retrieved cloud fraction and increases retrieved mean . As with all methods that seek to infer using passive radiometry, the presence of aerosol requires that threshold values be set in order to discriminate between cloudy and cloud-free columns. A technique for estimating thresholds for cloudy columns is discussed and demonstrated. Finally, it was found that surface type and mean inherent play major roles in defining retrieval accuracy.