Understanding the Role of Land-Atmospheric Coupling in Drought Forecast Skill for the 2011 and 2012 US Droughts

The overall goals of the proposed project is to understand land-atmospheric coupling processes in CFSv2 and their role in the predictability of drought development, intensification and termination, and to perform attribution and modeling studies for the improvement of drought predictions.

Background. In 2011 and 2012, the central US suffered intense drought. While the societal impact of these extreme events can be reduced through planning and preparation, the predictive skill of seasonal forecasts from models such as NCEP’s CFSv2 is low, which limits their practical use. This is particularly true during the North American summer, when the need for predictions is the greatest.

Recent research by PI Wood’s group has resulted in a better understanding of the role that land- atmosphere interactions play in drought predictability in seasonal forecast models. This work developed the Coupling Drought Index (CDI) to assess the representation of land-atmosphere feedbacks in forecasts. The research has shown that the hindcast climatological CDI in CFSv2 quickly deviates from the reanalysis-based CDI into a wetter state and demonstrated that land-atmospheric coupling breaks down in CFSv2 during drought conditions (dry coupling) leading to the weakening and premature termination of the drought conditions. The loss of seasonal forecast drought skill is attributed to the failure of CFSv2 to hold drought conditions, especially in the major droughts of 2011 and 2012.

Preliminary analyses indicate that increased (anomalous) terrestrial evapotranspiration in CFSv2 is leading to its inability to hold drought conditions. One hypothesis is that deep soil water is accessed in the Noah land model and evaporated to control a warm bias while an alternative hypothesis is that the increase in evapotranspiration is due to a lack of dynamic vegetation in the model, which allows for continued transpiration during a drought event (due to the use of a vegetation phenology based on climatology)

Summary of Proposed Work: To address the goal of the proposed project, a combination of historical (realtime) and prescribed CFSv2 forecasts will be used to compare with verification data (CFSR/CDAS) to analyze the local feedback mechanism, the large scale circulation and their interactions in the development, intensification and termination of the 2011 and 2012 droughts in North America. Specifically,
1. An analysis of the CFSv2 forecasts (leads out to September) made from April through June for each event. The CFSv2 ensemble forecast data will be used to compute time series of the coupling states and CDI following the approach of Roundy et al., 2013a, b). The CFSv2 skill will be compared to a benchmark based on a CDI-based Statistical Drought Forecast Model.
2. Recycling analysis to track the moisture sources of CFSv2 anomalous precipitation, to assess if the moisture is from local sources (anomalous ET in CFSv2) or from large scale advection.
3. CFSv2 ensemble reforecast experiments for the droughts of 2011 and 2012 to examine the role of vegetation parameterization in CFSv2 (Noah). The experiments will use real-time vegetation fraction observations and an advanced Noah land model with Multiple Parameterization (Noah-MP) that includes both fixed and dynamic vegetation options.

Relevance to the Program. The research directly addresses the needs identified in the call: “Proposals (that) will examine processes controlling drought development, intensification, and termination with a focus on predictability. Specifically, (work that) consider mechanistic studies involving model simulations and predictions to examine processes such as the role of land surface conditions, … and atmospheric feedbacks..”. The significance of the research is that the work will help understand the processes that lead to premature termination of drought in CFSv2 forecasts and its low forecast skill. Based on the anticipated results, the project can identify potential CFS model improvements.