As general circulation models (GCMs) increase in complexity, the number of physical processes representing the climate system increases. A central goal of climate science is to understand how uncertainty in these physical processes, translates into uncertainty in the system response. But the very complexity of the GCMs generates a troublesome question: how do you properly define a stable and unchanging reference system with which to compare constituent elements and characterize their uncertainties? To navigate around this, we introduce a simple energy balance model (EBM) that accounts for the transport of both sensible and latent heat in the atmosphere. We demonstrate that the EBM accounts for approximately 90% of the inter-model spread in the temperature response for an ensemble of GCMs subjected to increasing CO2. We then calculate the unique model patterns of ocean heat uptake, radiative forcing, and radiative feedbacks and demonstrate that 100-years after an abrupt quadrupling of CO2 above preindustrial values, the largest source of uncertainty is the pattern and amplitude of radiative feedbacks. Using this, we identify regions of high feedback uncertainty and show how this can bias temperature responses in other regions.

From November 2015 through March 2016, the Olympic Mountains Experiment (OLYMPEX) field campaign was conducted on the Olympic Peninsula to study how wintertime precipitation is modified as it passes over coastal mountains and to validate satellite-derived precipitation measurements from the U.S.-Japan Global Precipitation Measurement (GPM) mission. This project uses OLYMPEX data to explore the topographic effect on these Pacific frontal systems by examining cloud and precipitation structure on the leeward side (usually northeast) of the Olympic Mountains, where there is typically a minimum in precipitation relative to the windward side. While most research has focused on the structure of the windward side, this study uniquely examines the leeward side of the mountains. A radar managed by Environment and Climate Change Canada on Vancouver Island provided measurements of the vertical structure of the frontal systems over the northern Olympic Mountains, including intensity inferred from radar reflectivity. Using environmental data from NCEP North American Regional Reanalysis on the windward side (usually southwest) of the mountains, this study classifies the leeward radar data based on upstream synoptic conditions. It then examines the cloud and precipitation structure on the leeward side of the mountains in various atmospheric environments. Preliminary findings of this study reveal, for example, that the intensity of cloud systems on the leeward side is dependent on upstream stability. Locally, these findings will inform studies of snowpack and water supply as many reservoirs depend on precipitation that occurs on the leeward side. Outside of the Pacific Northwest, these findings can be applied to other midlatitude coastal mountain ranges on the west side of continents around the world.

Our research seeks to more accurately define how the Pacific Northwest regional climate will look in fifty to one hundred years. Our results help quantify the uncertainty in our climate predictions, a major shortcoming in climate research. Our research is interesting because it utilizes a high resolution numerical weather prediction model to examine regional climate change here in the Pacific Northwest. We are creating a large set (an ensemble) of predictions of climate change in the Pacific Northwest. The tool used in our research is the Weather Research and Forecast (WRF) model. WRF requires General Circulation Model (GCM, also known as Global Climate Model) data to feed it information on the larger-scale climate. A big part of our project is collecting GCM data from a variety of research centers worldwide, and preparing it to be used as input data for WRF. We expect to complete our model simulations by the end of Spring Quarter 2018. These results could be of direct use to a number of local agencies and policy makers so that they can shape their decisions around the future changes that our local climate may experience. We also expect our data to be of great use for further research in quantifying climate prediction uncertainty.

Aerosol particles, solid or liquid particles suspended in the air, can affect Earth’s climate by scattering and absorbing sunlight. PM_2.5, particles less than 2.5 micrometers in diameter, are largely responsible for degraded visibility and easily traverse respiratory system pathways causing adverse health effects in humans. Often, a majority of PM_2.5 is composed of organic matter formed through photo-chemical reactions of hundreds of volatile organic compounds (VOC). This process depends on nitrogen oxide radical (NO_X = NO + NO₂) concentrations in the atmosphere, now predominantly contributed by anthropogenic emissions. NO_X and VOC react to produce organic nitrates, thought to be significant contributors to PM_2.5, but for which measurements are generally lacking. My project involves developing a concrete assessment of the role of organic nitrates as contributors to PM_2.5. I am developing an analytical method to thermally desorb organic nitrates from atmospheric particles collected on filters and promptly decompose them into nitrogen dioxide (NO₂). The resulting NO₂ is measured with a Cavity Attenuated Phase Shift (CAPS) spectroscopy instrument. I am establishing whether there is a direct correlation between the concentration of organic nitrates in PM_2.5 and desorbed NO₂. The goal is to quantify organic nitrates without individually measuring each of the likely hundreds present in the ambient atmosphere. Moreover, the thermal desorption process provides information on the physical properties of organic nitrates, such as effective saturation vapor pressure, needed for air quality computer models to accurately simulate their contribution to PM_2.5. As a result, I will help develop a better understanding of how natural and anthropogenic emissions affect the composition of the atmosphere by allowing assessments of how PM_2.5 levels have changed in response to NO_X emission reductions by the Clean Air Act.

Wintertime air pollutant emission trends are poorly understood in comparison to emission trends in the summer mostly due to a lack of available and analyzed data on precursor air pollutants in the wintertime. This creates uncertainty in current atmospheric chemistry models. “WINTER 2015” was an aircraft research campaign launched in February 2015 that was executed to measure pollutant concentrations in addition to other data. All of which were relevant to the transportation, distribution, and chemical kinetics of pollutants and aerosols across the northeastern United States. The colder temperatures as well as lack of available sunlight have significant effects on chemical kinetics in the atmosphere. In particular, during the daytime, nitrogen oxides tend to produce ozone (O₃), a criteria pollutant regulated by the U.S. EPA, but nitrogen oxides destroy ozone at night. Longer nights, and lower sunlight during the day may mean that nitrogen oxides net destroy ozone for much of the winter. Current air quality models do not accurately simulate the nighttime chemistry of nitrogen oxides. I am examining the wintertime relationship between nitrogen oxides and O₃ using data taken in the WINTER 2015 campaign to gain insights into the extent to which wintertime nitrogen oxide emissions produce or destroy ozone. This information will then improve atmospheric chemistry models that are used to forecast air quality in various regions. It is expected that in the wintertime, the amount of reactive nitrogen oxide is limited relative to hydrocarbons. This nitrogen limited regime would mean that reactive nitrogen oxides have a greater role in the destruction of ozone, than the production of ozone.

Central Argentina in the lee of the Andes Mountains experiences some of the most intense thunderstorms on earth. These storms produce large hail, flash floods, and tornadoes, and adversely affect the people living in this region. The initiation, growth and hail production processes in these storms are poorly understood due to the lack of sufficient observations in this region. Currently, the only way to study the climatology of these events is through satellite remote sensing observations. This project examines spaceborne radar data from the Precipitation Radar (PR) on the Tropical Rainfall Measuring Mission (TRMM) satellite for the November and December period for the years 1998-2013. The purpose of this project is to better understand the temporal and spatial evolution of these convective systems near the Sierras de Córdoba in the lee of the Andes. The PR radar observations from TRMM were classified into 4 categories: 1) deep convective cores; 2) wide convective cores; 3) intermediate convective cores; and 4) broad stratiform regions. These categories are based on the height, area, and reflectivity of the storms. Reflectivity is the signal reflected back to the radar by the hydrometeors. It was found that initiation and rapid upscale growth occur in the vicinity of the Sierras de Córdoba. These results are crucial for the planning and execution of the upcoming field campaign in Argentina called RELAMPAGO (Remote sensing of Electrification, Lightning, And Mesoscale/microscale Processes with Adaptive Ground Observations). Knowledge about how convective storms behave in Argentina applies to other regions, such as the lee of the Rocky Mountain in the United States. Another benefit of this study is that legacy computer code that was written in the IDL programming language is now updated to a more structured format with Python making it easily used for future research.

People all over the world use forecasts from various sources such as the National Weather Service, local TV/radio stations, or apps on smartphones for a variety of applications. Examples include preparing for routine activities, such as planning outdoor events, or preparing for rare, yet hazardous scenarios like a thunderstorm passage near a sporting event. The most common quantities of interest are maximum/minimum temperatures, wind speed, and rainfall (chance and amount). To approximate these quantities, weather information sources utilize forecasts made from numerical weather prediction models of which there are about ten used in the United States. Numerical models of the atmosphere consist of equations that describe the current state of the atmosphere and how it changes with time and location. These equations are solved on powerful computers as the number of calculations are immense. Using the information from these models, individual forecasts can be made. Atmospheric Science students at UW are practicing forecast techniques through participation in a national competition called WxChallenge—a contest where participants from various academic institutions predict these quantities for selected US cities. For help with forecasting, the UW team has developed a website that holds a suite of model information which easily analyzes and compares that data and assesses the skill of each individual model. This website, however, needs updates and improvements. Therefore, I am converting the existing system to an object-oriented format by creating Forecast objects for individual weather models. For example, I have written a Forecast object for the DarkSky weather model using Python code. This type of improvement will reduce redundant code and allow for future developments to be implemented with ease. I am confident that this updated system will help the UW team improve the accuracy of their forecasts.