A study is conducted to determine the value of initial high resolution moisture fields on short-range precipitation forecasts in order to facilitate the design and effectiveness of integrated observing systems. A series of sensitivity experiments are performed with an idealized extratropical cyclone model to isolate impacts of initial mesoscale structure in the atmospheric water vapor on precipitation forecasts. The experiments are made with a 30 km version of the University of Wisconsin Non-Hydrostatic Model System (UW-NMS) in a baroclinic channel with zonal cyclic boundary conditions. The model is initialized for a rapid-growing idealized middle-latitude baroclinic wave possessing a 4,000 km wavelength. The idealized cyclone solution rapidly amplifies during the first hours and is characterized by a robust occluded structure by 36 hours. Precipitation develops along the frontal regions and in the warm sector; however, it is most pronounced in the vicinity of the triple point.
Several baseline control case conditions are used to give some generality to the idealized cyclone model. Two conditions for tropospheric static stability are considered: stable and less stable atmospheres with temperature lapse rates of 2.5 degrees Centigrade and 6.5 degrees Centigrade per km,respectively. In addition, numerical model versions with and without moist convective parameterization are used.
Thirty-six hour model simulations are compared for cases with and without an initial lower tropospheric mesoscale water vapor anomaly. The initial moisture anomaly has a small circular region (diameter of 400 km) where moisture is enhanced by 3 g/kg surrounded by a larger region of slightly reduced water vapor so that the volume-total water vapor content is not changed. A small corresponding temperature anomaly is introduced to conserve the unperturbed virtual temperature field and thus preserve the prescribed mass distribution. The moisture anomaly is located in the warm sector of the developing cyclone such that it is advected into the region where precipitation develops in the control simulation by 12 hours. Several experiments, each with slightly different locations of the anomaly, are made for each baseline model control in order to determine the robustness and properties of the anomaly response.
The moisture anomalies affected the precipitation for the cases where the moisture anomaly was advected into the regions where precipitation developed in the baseline control simulation. The effect was most pronounced in regions where there were prefered modes of growth for mesoscale and convective-scale systems. The overall tropospheric static stability has an important influence on the nature of such growing modes. The results demonstrated a positive impact of initial mesoscale moisture information during the first 24 hours of a mesoscale model forecast in selected parts of the domain and provide examples of where and when it may be most important to attempt retrieval of water vapor information from high resolution satellite and other remote sensing systems.