North American Monsoon (NAM) thunderstorms in the Southwest United States account for nearly 50% of the annual precipitation in this region, yet these phenomena have been relatively understudied and are difficult to predict. The NAM brings a host of hazards in a part of the country experiencing very rapid population growth: life-threatening flash floods, damaging microburst winds, sudden mile-high dust storms that sometimes form in association with thunderstorm outflows, and lightning-triggered wildfires made even more dangerous by the outflows (lightning is responsible for 62% of the wildfires in central Arizona). The timing, location, and intensity of such hazards are challenging to predict, even with state-of-the-art numerical weather prediction (NWP) models having sufficient resolution to explicitly resolve storms. A primary deficiency is that observations of the atmosphere above the surface used to initialize NWP models are currently too widely distributed compared to the spatial scale of thunderstorms; also, these data are available just twice daily, whereas a typical thunderstorm has a lifetime of less than one hour. Therefore, the goal of this project is to determine the optimal distribution and types of much-improved arrays of instruments within both future NSF-supported field campaigns and for a future statewide operational high-resolution ?mesonetwork? in Arizona to have the greatest positive impacts on the predictability of the initiation, evolution, and upscale growth of thunderstorms in the NAM. The findings from the study should provide the many current stakeholders of the University of Arizona NWP system with information about the expected value (and cost) of a statewide mesonet for improving the prediction of NAM weather. The project will also inform decisions regarding the optimum instrument deployment strategies to be made for future higher-resolution field campaigns designed to improve understanding and predictability of storms in complex terrain. Lastly, an important result of the effort will be the development of the modeling and data assimilation infrastructure needed to obtain four-dimensional consistent datasets of temperature, moisture, winds and precipitation using the optimally-determined arrays of observations in the future. The project will use a novel application of the Observing System Simulation Experiment (OSSE) methodology to determine the optimal configurations for a future operational Arizona state mesonetwork and the complementary requirements for the design of more densely spaced instrument arrays in future mesoscale field campaigns. OSSE is a modeling experiment used to evaluate the value of observing system when actual observational data are not available. Each new (not currently operational) instrument type can be introduced, along with appropriate error variances, in a systematic manner by using Ensemble Kalman Filter (EnKF) data assimilation to evaluate relative impacts on model predictions. The innovative OSSE approach for optimizing network design has the potential for high reward as it represents a fundamentally different approach from what has been previously used for state-operated mesonet design considerations and large field campaigns, thus making such decisions more cost-effective. The research team has ample peer-reviewed experience conducting OSSEs for the following synthetic observations to be investigated: GPS vertically integrated precipitable water vapor, vertically-resolved measurements of water vapor from MicroPulse Differential (MPD) absorption lidars, winds from Doppler Lidars, and data from rotary-wing Uncrewed Aircraft Systems (UAS) data and 3-hourly soundings. The OSSEs will be conducted within the framework of the University of Arizona WRF modeling system and the ensemble adjustment EnKF available within NCAR?s Data Assimilation Research Testbed (DART). An important benefit of the research is development of the scientific and technical infrastructure needed to create Four Dimensional Dynamically Consistent (4DDC) datasets from the assimilation of the various observing system data, since many of the governing factors in performing the data assimilation will have been addressed during this research. Because the project will develop the 4DDC infrastructure prior to commencement of any future field campaign, scientists will be able to utilize the field 4DDC datasets in their research more efficiently and quickly. Thus, the project represents both a risk reduction effort regarding optimization of network design and the means by which greater use of the data can occur. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.