In this paper, we examine for a sample of Los Angeles residents their willingness to pay to prevent significant climate change. We employ a fractional factorial design in which various climate change scenarios differing in ways consistent with existing variation in climate are presented to respondents. These are contrasted to respondents’ current climate before willingness to pay is elicited. Thus, the focus is on climate change as it may be experienced locally. We also try to determine the kinds of value that are driving respondents’ concerns. Among the key findings are that for these respondents, climate change leading to warmer local temperatures is a greater worry than climate change leading to colder local temperatures. In addition, climate change leading to less precipitation locally is of more concern that climate change leading to more precipitation locally. Finally, use value may be the most important kind of value, but a more cautious interpretation is that respondents are not yet able to clearly distinguish between different climate change consequences.

The Santa Ana winds represent a high-impact weather event owing to the intimate relationship between the extremely dry, fast winds and the wildfire threat. The winds can be locally gusty, particularly in the complex terrain of San Diego county, where the airflow has characteristics of downslope windstorms. These winds can cause and/or rapidly spread wildfires, the threat of which is particularly acute during the autumn season before the onset of winter rains. It remains a day-to-day challenge to accurately predict wind gust speed, especially in the mountainous regions.

Our study employs large physics ensembles composed of high-resolution simulations of severe downslope windstorms that involve an exhaustive examination of available model physical parameterizations. Model results are calibrated and validated against the San Diego Gas and Electric (SDG&E) mesonet observations, a dense, homogenous, and well-positioned network with uniform high quality. Results demonstrate model horizontal resolution, model physics, random perturbations and landuse database can have a material effect on the strength, location and timing of Santa Ana winds in real-data simulations. A large model physics ensemble reveals the land surface model to be most crucial in skillful wind predictions, which are particularly sensitive to the surface roughness length. A surprisingly simple gust parameterization is proposed for the San Diego network, based on the discovery that this homogeneous mesonet has a nearly invariant network-averaged gust factor. The gust forecast technique is of special interest in the context of routine weather combined with atmospheric humidity and fuel moisture information.

A real-time wildfire threat warning system, the Santa Ana Wildfire Threat Index (SAWTI), has been developed to effectively communicate the upcoming Santa Ana wind strength with respect to the anticipated fire danger to first responders and the public. In addition to the wind and gust forecast techniques, attempts have been made to skillfully model two essential elements that SAWTI is heavily dependent on, i.e., the live fuel moisture and the greenness, using meteorological information. The models developed can skillfully determine these essential elements from both forecast and reanalysis data.

A model to predict the probability of large hail in a thunderstorm using pre-storm environmental parameters is developed with an ensemble decision tree method known as Random Forest. Variables integral to hail forecasting are investigated and selected for inclusion in the model. Using a predefined database of observed atmospheric soundings, the model is tested and found to have positive skill, similar to current operational techniques. Applying the model to North American Regional Reanalysis (NARR) data, a recent climatology of hail threat in the central United States is created and compared to observations. Model statistics, generated by defining probability thresholds to detect large hail and comparing the results to observed hail reports, show that the model has an advantage over other techniques in correctly forecasting hail occurrence. Trends in the statistics are explored and found to have a dependence consistent with spatial and temporal variability in atmospheric variables over the US, and improvements to the model based on these findings are proposed.

Predecessor rain events (PREs) are coherent mesoscale rainstorms that occur well in advance of recurving tropical cyclones (TCs) and have a high potential to cause flooding and adverse societal impacts. In 2007, a PRE associated with TC Erin produced record-breaking rainfall (> 350 mm) across southern Minnesota and Wisconsin, contributing to seven fatalities and over $170 million in property damage. A series of idealized numerical simulations is conducted using an aquaplanet version of the Weather Research and Forecasting model Advanced Research core (WRF-ARW) v.3.6 to examine the influence of cloud-radiative processes on the development of PREs.

This study finds that cloud-radiative feedback (CRF), the interaction of hydrometeor cloud species with longwave (LW) and shortwave (SW) radiation, produces a more robust PRE structure with stronger convective activity and, ultimately, more precipitation. It is demonstrated that LW cooling associated with clouds outside of the PRE region induces a stronger horizontal pressure gradient that enhances low level con ergence and drives more vigorous ascent. Therefore, the primary radiation driver of PRE formation occurs outside of the PRE itself and is based on how the model responds radiatively to low clouds. In addition, the warming component of CRF enhances parcel buoyancy and reduces vertical stabilities within the PRE structure. In particular, the distribution of cloud ice produces a greater depth of in-cloud, primarily LW warming conducive to stronger convective processes. In contrast, SW CRF effects weaken PRE development through a combination of in-cloud cooling and low level cloud top warming, countering the LW CRF effects responsible for greater rainfall production. Moreover, PRE formation is found to be sensitive to the diurnal cycle, resulting in faster development at night and slower development during the day. A seasonal sensitivity also brings about weaker PRE intensification during the summer season; shorter nights and longer days minimize favorable LW CRF effects and maximize unfavorable SW CRF effects.

PRE development depends on the balance between subtle, competing processes involving radiation and microphysics schemes. The high sensitivity of PRE initiation and intensification to variations in CRF cannot be ignored and necessitates further research and observations to improve operational forecasts. Because the recipe for PRE development requires many individual ingredients to come together correctly, accurately predicting heavy rainfall events associated with tropical cyclones remains a daunting forecast challenge.

Despite our increased understanding in the relevant physical processes, forecasting radiative cold pools and their associated meteorological phenomena (e.g. fog and freezing rain) remains a challenging problem in mesoscale models. In this thesis we recognize that our current modeling approach is flawed and incomplete, rendering us unable to forecast these high impact events. Using the Weather Research and Forecasting (WRF) model, fundamental deficiencies are exploited and new physical parameterizations are introduced to address these issues without degrading the model in other areas and time.

It was found that the default model diffusion, which is calculated on model sigma coordinates, in addition to the 6th order numerical filter, prevented the formation of cold pools. Furthermore, soil moisture in and around valleys from both natural and anthropogenic sources, the vertical resolution, and model physics were all found to be important and play at least some role in forecasting these events. Finally, key physical processes governing the evolution and life cycle of cold pools were missing and subsequently introduced to a boundary layer parameterization to substantially improve its forecasting ability. Through the knowledge presented here, this and other key modifications have accomplished a numerical weather model of more general use and applicability which will not only be of use to the modeling community but society at large.

Using the Weather Research and Forecasting (WRF) model, an intra-model physics ensemble was created for July 2005 in hopes of finding a physics configuration that could most accurately recreate the typical summertime conditions of Southern California. The ensemble consisted of twenty-four members that mixed different Land Surface Models (LSM) with various Planetary Boundary Layer (PBL) parameterizations currently offered in WRF. It was found that the Pleim-Xiu LSM coupled with either the asymmetric convective model version 2 or the Yonsei University PBL scheme created the most skillful ensemble member.

In addition to the physics ensemble, short integrations along with different initializations proved to be beneficial. Initializing the model off of the North American Regional Reanalysis (NARR) instead of the North American Mesoscale (NAM) model was, in this case, extremely useful. The most accurate simulation of July 2005 was found when taking integration methods, initializations, and model physics into consideration.