You are here

Resources

PhD Seminar - Beth Sims

Fri, 2017/03/31 - 11:00am
Seminar
Presenter: 
Elizabeth Sims
Location: 
353 LOV

Event Details

Title: IMPROVING SATELLITE-BASED SNOWFALL ESTIMATION: A NEW METHOD FOR CLASSIFYING PRECIPITATION PHASE AND ESTIMATING SNOWFALL RATE. Abstract: In order to study the impact of climate change on the Earth’s hydrologic cycle, global information about snowfall is needed. To achieve global measurements of snowfall over both land and ocean, satellites are necessary. While satellites provide the best option for making measurements on a global scale, the task of estimating snowfall rate from these measurements is a complex problem. Satellite-based radar, for example, measures effective radar reflectivity, Ze, which can be converted to snowfall rate, S, via a Ze relation. Choosing the appropriate Ze-S relation to apply is a complicated problem, however, because quantities such as particle shape, size distribution, and terminal velocity are often unknown, and these quantities directly affect the Ze-S relation. Additionally, it is important to correctly classify the phase of precipitation. A misclassification can result in order-of-magnitude errors in the estimated precipitation rate.Using global ground-based observations over multiple years, the influence of different geophysical parameters on precipitation phase is investigated, with the goal of obtaining an improved method for determining precipitation phase. The parameters studied are near-surface air temperature, atmospheric moisture, low-level vertical temperature lapse rate, surface skin temperature, surface pressure, and land cover type. To combine the effects of temperature and moisture, wet-bulb temperature, instead of air temperature, is used as a key parameter for separating solid and liquid precipitation. Results show that in addition to wet-bulb temperature, vertical temperature lapse rate also affects the precipitation phase. For example, at a near-surface wet-bulb temperature of 0°C, a lapse rate of 6°C km-1 results in an 86 percent conditional probability of solid precipitation, while a lapse rate of -2°C km-1 results in a 45 percent probability. For near-surface wet-bulb temperatures less than 0°C, skin temperature affects precipitation phase, although the effect appears to be minor. Results also show that surface pressure appears to influence precipitation phase in some cases, however, this dependence is not clear on a global scale. Land cover type does not appear to affect precipitation phase. Based on these findings, a parameterization scheme has been developed that accepts available meteorological data as input, and returns the conditional probability of solid precipitation.Ze-S relations for various particle shapes, size distributions, and terminal velocities have been developed as part of this research. These Ze-SRelations have been applied to radar reflectivity data from the CloudSat Cloud Profiling Radar to calculate the annual mean snowfall rate. The calculated snowfall rates are then compared to surface observations of snowfall. An effort to determine which particle shape best represents the type of snow falling in various locations across the United States has been made. An optimized Ze-S relation has been developed, which combines multiple Ze-S relations in order to minimize error when compared to the surface snowfall observations. Additionally, the resulting surface snowfall rate is compared with the CloudSat standard product for snowfall rate.