A regional atmospheric model, the Weather Research and Forecasting (WRF) model, will be used to improve short and long-term forecasting capability for the North American Monsoon System. This study is motivated, in part, by the North American Monsoon Experiment (NAME), which has the objectives of understanding and predicting warm season convective processes in complex terrain and determining the response of warm season precipitation to slowly varying oceanic and continental surface conditions. In addition, physical understanding of how the North American Monsoon may be affected by anthropogenic climate change is an urgent question in the water-scarce Southwest United States and northwestern Mexico.
The scientific objectives of this project are: to explain the physical mechanisms which lead to monsoon rainfall and determine the sensitivity of numerical weather forecasts to data specification; to assess the potential for improvement in seasonal climate prediction in North America for the warm season; and to assess the potential effects of anthropogenic climate change on the core monsoon region. These objectives will be addressed using WRF as a numerical weather prediction and regional climate model and, in the process, developing model applications, such as an adjoint modeling system and spectral nudging capability.
Broader impacts are in the societal value of improved forecasts of the North American Monsoon and projections of how it may change in the future. This work will address the identified priorities of the weather and climate community and societal needs, as identified through collaborative links the PI maintains with these stakeholders. The research and teaching responsibilities of the PI will facilitate multidisciplinary interactions (including the development of a new hydrometeorology program) and cooperation with Mexico. The project will provide an example of Hispanic achievement and mentorship in the physical sciences.
BACKGROUND The North American monsoon is the period of late summer precipitation that occurs in northwest Mexico and the United States. In contrast to the winter, when large-scale mid-latitude cyclones produce geographically widespread and steady precipitation, monsoon precipitation is in the form of localized convective thunderstorms. An improved ability to forecast and project the monsoon, considering timescales of short term weather prediction (out to few days), warm season seasonal forecasts (out to few months), and climate change projection (over the coming century), was identified as a major research challenge by the North American Monsoon Experiment (NAME). This research project developed the use of a regional atmospheric model, specifically the Weather Research and Forecasting model, to address this broad issue. Therein, the specific research questions in this project were: 1) Can North American monsoon thunderstorms be explicitly represented at a convective-resolving (less than 2km) resolution and can the model forecast sensitivity to the specification of initial data be assessed? 2) Can a regional climate model improve seasonal forecasts for the North American monsoon? and 3) How is the overall character of North American monsoon climate changing in association with anthopogenically-forced global climate change, both in the recent observational record (over the past 60-years) and in the coming century? PRIMARY OUTCOMES The project considered use of the Weather Research and Forecasting model as both a numerical weather prediction (NWP) model and regional climate model, as well as various sources of gridded climate data. Considering short-term weather forecast type simulations of Intensive Observing Periods (IOPs) during NAME, it was demonstrated that: 1) use of a convective-resolving grid spacing (of approximately several kilometers) is required to reasonably represent severe weather that occurs in with association organized convection and 2) forecasting this severe weather is very sensitive to the specification of initial data in areas where thunderstorms initiate, such as the Sierra Madre Occidental in Mexico, or surges of moisture from the Gulf of California. A climatology of retrospective warm season seasonal forecasts for North American were generated for the period 1981-2000, by dynamically downscaling data from the Climate Forecast System Model (CFS), the operational global seasonal forecast model used by NOAA’s Climate Prediction Center. Analysis of these hindcast simulations showed that a regional climate model can slightly improve warm season seasonal forecasts, but ultimately depends on how large-scale atmospheric circulation variability can be deterministically represented in CFS. Several "well performing" global climate change projection models from the Couple Model Intercomparison Project, Version 3 (CMIP3) were also dynamically downscaled, and results of these simulations were compared to observed long-term trends in warm season climate in the recent historical record. It was generally found that the North American monsoon is changing in way that comports with a "wet-gets-wetter, dry-gets-drier" hypothesis, such that transitions in monsoon precipitation are becoming more extreme in association with both the climatology and natural climate variability. Project results have been disseminated as four professional journal articles, either published or pending at the project completion date, as well as numerous professional conferences. BROADER IMPACTS Project findings substantively advanced scientific capability to improve North American monsoon forecasts and projections. The modeling experiments considering short-term simulations of NAME IOPs provide guidance for the design of a high resolution operational model forecast system and indicate "hot spot" locations where future collection of in-situ data in needed in a permanent monsoon observing system in Mexico. A follow-on NSF-funded study is currently taking place which will evaluate the effect of assimilating GPS-based atmospheric precipitable water measurements from a pilot array in northwest Mexico during the 2013 monsoon, which should eventually dovetail into the establishment of a new permanent observing system of GPS meteorological instruments in Mexico. Conclusions on how dynamical downscaling can potentially improve warm season seasonal forecast have been communicated directly to NOAA’s Climate Prediction Center, and provide motivation for use of a multiple regional model, ensemble approach to improve operational warm season climate forecasts in the near future. Climate change projections of the monsoon have been actively communicated to various natural resource stakeholders in the Southwest U.S. and Mexico during the course of the project, through various local workshops and professional exchanges. The regional climate model data generated within the project are currently being used for water resource assessment and to project how severe monsoon weather will change in the future. The project has contributed to the graduate education of four students in the Department of Atmospheric Sciences at the University of Arizona (UA), two of which successfully attained M.S. degrees. It has also supported the professional development of a postdoctoral researcher. Through collaboration with UA Laboratory for Tree Ring Research, UA Department of Hydrology, and National Center for Atmospheric Research, and NOAA's Climate Prediction Center, the project work has resulted in synergistic, multi-disciplinary research deliverables to address climate issues in the Southwest U.S. and beyond.
