Spring 2016 Abstracts
Jan 29, 2016
Extreme precipitation events are drawing more and more attention under the background of global climate changing. They make a huge impact on our economy and daily life and more evidences showing extreme precipitation events are on the rise. However nowadays, the performance of GCMs to forecast the extreme precipitation events are still relatively poor.In our study, we employed the CWRF (Climate extension of the Weather Research and Forecasting model) to downscale the output from ERA-interm in order to simulate the extreme precipitation events during 1980-2010. We further incorporate a comprehensive ensemble of multiple alternative representations for major physical processes, ranges from land–atmosphere–ocean, convection–microphysics-precipitation to cloud–aerosol–radiation.
Our results demonstrate: generally speaking, the ensemble CWRF system enhances the skill of precipitation significantly. It also helps us to find out the optimized physics ensemble combination to improve extreme precipitation prediction over the U.S wth CWRF. It will illustrate to some aspects of extrem events, whether and how should we adapt an ensemble method or improve and adapt a more physically meaningful scheme.
Keywords: CWRF, ensemble simulation, extreme precipitation events, dry spells, droughts, heavy rainfalls, and floods
Feb 02, 2016
The NASA Studies of Emissions and Atmospheric Composition, Clouds, and Climate Coupling by Regional Surveys (SEAC4RS) field campaign occurred in the south central U.S. during August and September of 2013 with a complement of surface, aircraft, and satellite measurements to further understand atmospheric chemistry in the troposphere and stratosphere. Flights on the NASA DC8 and ER2 from Houston were augmented by a network of six coordinated ozonesonde launching stations (SEACIONS) that included St. Louis, Houston, and Huntsville. Stratospheric intrusion events were studied to investigate the impact on free troposphere and boundary layer ozone. Relative humidity profiles from the AIRS (Atmospheric InfraRed Sounder) satellite instrument, confirmed the presence of stratospheric intrusions when dry stratospheric air was co-located with ozone-rich, nitrogen dioxide deficient, and carbon monoxide deficient air as measured by the ozonesondes and aircraft instruments. The Goddard Earth Observing System Model Version 5 (GEOS-5) was run with stratospheric tracers to determine the quantitative impact of stratosphere-to-troposphere transport (STT) on the ozone budget. The CMAQ model was simulated to evaluate its skill for the stratospheric intrusion events. Case studies through the campaign are presented to show the variability and impact of the stratospheric intrusions that occurred. Overall stratospheric intrusion events during this period increased free tropospheric ozone by 20-40 ppb, including just above the boundary layer. However, they generally had no noticeable effect on the boundary layer composition.
Feb 09, 2016
Rapid changes in the Earth’s ice sheets, glaciers, and sea ice in recent decades require precise measurements to understand their driving mechanisms. Precise calibration of instruments used to measure these surfaces, such as photon-counting laser altimeters, is needed in order to measure optical properties of snow covered ice sheets and sea ice using subsurface scattered photons.
Photons traveling into snow, ice, or water before scattering back to an altimeter receiving system (subsurface photons) travel farther and longer than photons scattering off the surface only, causing a bias in the measured elevation. I seek to identify subsurface photons in a laboratory setting using a flight-tested laser altimeter (MABEL – Multiple Altimeter Beam Experimental Lidar) and to quantify their effect on surface elevation estimates for laser altimeter systems. I will compare these estimates with previous laboratory measurements of green laser light transmission through snow, as well as Monte Carlo simulations of backscattered photons from snow, and explore the potential to measure snow grain size using this understanding of subsurface scattered photons.
Feb 16, 2016
Non-dispersive infrared (NDIR) sensors are a low-cost way to observe carbon dioxide concentrations in ambient air, but their specified accuracy and precision is not sufficient for scientific applications. An initial evaluation of the SenseAir K30 sensor in a lab setting showed that when accounting for individual offset bias, the sensor has an error of approximately five parts per million when compared to a Los Gatos Research greenhouse gas analyzer that uses cavity ring down spectrometry. Through further evaluation, after correcting for environmental variables with coefficients determined through a multivariate linear regression analysis, the calculated difference between the NDIR sensor and the higher-accuracy instrument had a standard deviation in the 2 to 3 ppm range. Thus, after individual correction and calibration, the K30 sensor has the potential to compliment traditional higher accuracy but high cost instruments in certain observation applications.
Feb 23, 2016
Multiple causes have been put forward to explain Arctic sea ice loss, including a decline of the multiyear sea ice cover, warming due to ocean and atmosphere heat transport, and enhanced absorption of solar radiation by the ocean. This presentation lays the groundwork for further investigations of the contributions by different terms in the Arctic energy budget equation to sea ice melt in various sectors of the Arctic Ocean and surrounding seas.
Mar 08, 2016
Parcel trajectory analysis is a powerful analytical tool for the atmospheric and oceanic sciences, with applications as wide-ranging as the analysis of three-dimensional convective updrafts in tropical cyclone simulations to the tracing of pollutants at the synoptic scale. When parcel trajectories are computed from model output wind data, computational stability requirements generally require use of a computational time step considerably smaller than the model output time step. The linear interpolation of model wind data to trajectory computational time steps can be a significant source of error, particularly when the model output data is available at less-than-ideal time resolution for the flow field under study. Recently, a trajectory computation code has been developed for supercell tornadogenesis modeling applications, one that iteratively solves for local advective flows and then accounts for the advection of the wind field between model output time steps, with the goal of reducing the time interpolation errors. Here, a trajectory computation routine is being developed for analysis of tropical cyclone simulation data, which also incorporates an advection correction procedure. This seminar will focus on the development of this new methodology, and also discuss preliminary results and how they may be valuable for examining processes critical to tropical cyclone intensification.
Mar 22, 2016
Deep convection acts to vent boundary layer air into the free troposphere, often removing pollutants from the surface and decreasing surface ozone concentration. These convective updrafts carry ozone and its precursors out of the boundary layer and into a regime with longer lifetimes and potential climate implications through positive radiative forcing. Depending on the extent of mixing and photochemical reactions within convective updrafts and downdrafts, cleaner air may be transported downward into the boundary layer. Additionally, convection is a significant mechanism for altering trace gas and meteorological profile shapes. Since realistic model profile shapes are essential for accurate satellite retrievals of trace gas column amounts, understanding the effects of deep convection on profile shapes and surface concentrations is critical for air quality modeling and forecasting. This presentation will focus on thunderstorms observed during NASA’s DISCOVER-AQ mission and their effects on air quality. Presented here are cases from the July 2011 Maryland and the July 2014 Colorado deployments, where deep convection either acted to temporarily reduce surface ozone concentrations, or resulted in a reduction for longer durations. Cloud-resolved simulations of a DISCOVER-AQ storm was conducted using the Weather Research and Forecasting (WRF) model at 1-km and 330-m horizontal resolutions to assess the influence of deep convection on surface composition and lower tropospheric vertical profiles. A lightning nudging technique was used to improve the timing and location of the simulated storms relative to the observed radar reflectivity. These simulations are evaluated through comparisons with radar, aircraft, tethered balloon, and surface observations collected during the DISCOVER-AQ deployments.
Mar 29, 2016
Reduced-form models, or simple climate models (SCMs), are a class of model used to understand the effects of anthropogenic perturbations on the climate system. SCMs are easy to use and computationally inexpensive, making them an ideal model for a variety of analyses and important for decision-making-related and scientific research. In this study we compare two SCMs, Hector v1.1 and MAGICC 5.3, to diagnose model behavior and understand fundamental responses of the climate system. Hector is a new reduced form climate carbon-cycle model, while MAGICC 5.3 is a well-known SCM, commonly used in the literature. Previous studies have noted the importance of investigating model behavior with the ultimate goal of understanding important indicators, such as changes in the carbon cycle or transient climate response. In this study, we discovered that Hector v1.1 responds differently from MAGICC 5.3 to stylized climate perturbations. For example, some non-linearities within Hector v1.1 were expected and resulted from improvements to the model’s carbon cycle representation. Still other differences were unexpected, such as a negative temperature response to a black carbon perturbation, and require more investigation. This initial study allowed us to identify differences within the two SCMs and understand fundamental model responses to anthropogenic perturbations. The next phase of our work seeks to clarify what role SLCF have in modifying the climate system by utilizing stylized experiments from CMIP5 with SCM emulations.
Mar 29, 2016
The initialization of a dynamically consistent storm-scale environment and the resulting forecast is one of the biggest challenges of convective-scale numerical weather prediction (NWP). Among the most critical sets of observations that can be used for initializing storm-scale NWP systems come from the United States Weather Surveillance Radar-1988 Doppler (WSR-88D) network. However, only the horizontal components of Doppler radial velocity (radial wind) observations are assimilated at a length scale ≥ 12-km. Some initial sensitivity experiments are performed to understand the radial wind assimilation procedure in the operational grid-point statistical interpolation (GSI) data assimilation system. These experiments are then tested in an operational configuration of the North American Mesoscale Rapid Refresh (NAMRR) forecast system during a heavy precipitation event 30-31 October 2015 using a hybrid three-dimensional variational-ensemble (3DEnVar) data assimilation algorithm. This seminar will discuss the results of the sensitivity experiments for addressing the assimilation length-scale and their effects on the case study analysis and forecast as well as some initial thoughts on addressing the assimilation of vertical velocity.
Apr 05, 2016
While major advances have been made in forecasting hurricane track, intensity forecasting has been a large focus of recent tropical cyclone research. Some hurricanes dissipate before becoming high intensity storms, and others develop into major hurricanes. Given the lack of available widespread high-resolution data, model simulations are a useful tool for investigating the inner core dynamics and thermodynamics of storms and how these factors relate to intensity changes. The Weather Research and Forecasting (WRF) ARW Model is used to examine the intensity and structural changes, including the Secondary Eyewall Formation, of Hurricane Earl, as well as the energetic changes associated with its Rapid Intensification (RI), which occurred beginning on 29 August and ending on 31 August, attaining a peak wind of 59.1 m s-1 and a minimum central pressure of 931 hPa. Model verification is performed in order to investigate the quality of the simulation relative to observations, in terms of track, intensity, and structure. Finally, a kinetic energy budget analysis is employed in order to better understand how energy changes occur, both spatially and temporally, within hurricanes undergoing RI, as energy exchanges between the storm system and the environment are expected to play an important role in the intensity changes of hurricanes.
Apr 12, 2016
Silvia Da Silva
The National Oceanic and Atmospheric Administration (NOAA) Unique Cross-track Infrared Sounder (CrIS)/Advanced Technology Microwave Sounder (ATMS) Processing System (NUCAPS) is the official NOAA system retrieving atmospheric vertical temperature and moisture profiles (AVTPs and AVMPs) from CrIS and ATMS measurements. Both state-of-art instruments are currently onboard the Suomi National Polar-orbiting Partnership (S-NPP) spacecraft, launched on October 28th 2011, as part of the Joint Polar Satellite System (JPSS). This study investigates the performance of atmospheric stability indices and parameters (SIPs) computed from NUCAPS AVTPs and AVMPs in order to verify their overall quality and applicability to the operational meteorological routine. The methodology considered comparisons between conventional and dedicated/reference radiosonde observations (RAOBs) with the closest NUCAPS retrievals and analysis profiles from the European Centre for Medium-Range Weather Forecasts (ECMWF) global model, collocated within a maximum radius of 50 km and ±1-hr time difference. Parameters evaluated include Total Precipitable Water, Lifted Index, K-Index, Total-Totals Index, as well as the recently developed Galvez-Davison Index, optimized for the tropics. The SIPs were computed from each of the three sources of soundings, NUCAPS, RAOBs, and ECMWF, and intercompared with proper metrics. Evaluation is divided by latitudinal bands, mid-latitudes (60N to 30N) and tropics (30N to 30S), covering a very comprehensive sample of RAOBs resulting in approximately 10000 for the mid-latitudes case, and ~3700 for the tropics. Results suggest that it is possible to capture synoptic-scale convective signatures, which would result in an additional and complementary information source for the weather forecasting. Comparison results over severe weather cases are also presented.
Apr 12, 2016
The urban heat island (UHI) effect can lead to serious health effects and social impacts in densely populated urban areas. When coupled to coastal cities like New York City (NYC), the complex interacting processes present a significant challenge to forecasting models in correctly predicting air temperatures and flow fields. This study uses the Weather Research and Forecasting (WRF) model and its available urban parameterizations to simulate the NYC extreme heat event of July 17-20, 2013. WRF output is compared to surface meteorological observations and analyzed to determine trends associated with identified UHI and land-sea breeze effects. Analysis results are presented, along with additional efforts associated with the overall research project.
May 03, 2016
Systematic model errors significantly contribute to the total forecast error, about 1/3 of the range of total errors, even in state-of-the-art weather prediction models like Global Forecast System (GFS). Our goals are to (1) estimate and reduce these systematic errors empirically and (2) provide guidance to monitor the impact of improved physical parameterizations. Danforth and Kalnay in three papers in 2007 and 2008 showed that the systematic components of the errors can be corrected by online model bias correction. This not only produced slightly better 5-day forecast error compared to the standard a posteriori statistical bias correction but also significantly improved the random errors.
We follow of Danforth and Kalnay except that we use a different approach to estimate the 6 hour forecast bias. To get the best estimate of the systematic errors before they start to grow non-linearly, we calculate average 6 hour analysis increment (AI), which is Analysis (A) –Background (B), an approach similar to Klinker and Sardeshmukh (1992). We conducted the analysis for 2012, 2013, and 2014 using data kindly provided by Dr. Fanglin Yang. We also analyzed Empirical Orthogonal Functions (EOFs) to find the leading components of the errors in the diurnal cycle. The results are extremely encouraging. The fields of mean AI for surface pressure, temperature, specific humidity, and winds at all model levels were quite similar over different years, indicating that the model bias is robust despite changes in model resolution and in data assimilation scheme. By comparing similar analysis at lower resolutions, we found that the 6hr model bias takes place on large scales and is not dominated by smaller scale phenomena. The analysis of errors in diurnal cycle shows that the leading 4 AI EOFs represent the diurnal cycle errors exceedingly well, and are also robust and stable. This implies the possibility of correcting the GFS diurnal errors with just 4 terms in the time derivative of each variable.
May 03, 2016
Can a weather balloon tell us something about the ocean? Can an ocean buoy tell us something about the atmosphere?
Coupled ocean-atmosphere models are increasingly important to improving numerical weather prediction for a range of phenomena including ENSO, monsoons, and tropical cyclones. Operational centers typically initialize the atmosphere and ocean components of these models with analyses from separate data assimilation (DA) systems in what is called weakly coupled DA. However, this separation can cause problems in subsequent forecasts. Strongly coupled DA, in comparison, handles the entire ocean-atmosphere system as a single state. This creates a more balanced analysis and extracts more information from observations by utilizing the cross-domain background-error covariance.
The Climate Forecasting System v2 Local Ensemble Transform Kalman Filter (CFSv2-LETKF) is being developed as a prototype system for the Indian Monsoon Mission Directorate. Initial results with observing system simulation experiments (OSSEs) show the benefits of strongly coupled DA. The system design presented should allow for straightforward implementation of operational strongly coupled DA for nearly any coupled model, including all subsystems of climate models (ocean, atmosphere, ice, land, surface waves, etc).
May 6, 2016
Elevated levels of tropospheric ozone caused by its precursor emissions of NOx and VOCs have a negative impact on human health and crops. Observations show surface ozone is enhanced at high temperatures due to a combination of meteorological and chemical factors. We investigate the representation of ozone precursors within the CMAQ (Community Multi-scale Air Quality) air quality model, with a focus on assessing how well the model represents NOx and HOx chemistry.
May 6, 2016
One-dimensional updraft models are at the heart of the mass flux approach to representing the effects of sub-grid scale convection in weather and climate models. While these models diagnose the mean properties of convection, it has been recognized in the modeling community that the variability of sub-grid scale convection also plays an important role and must be addressed. The only viable modeling approach at this time is the so-called multiple mass flux approach, whereby an ensemble of one-dimensional updraft models is used to characterize convective variability. While this framework has some advantages, it has practical and theoretical flaws. Here we introduce a statistical updraft model. While conventional one-dimensional updraft models diagnose only the mean heat, moisture, and momentum of a convective updraft ensemble, our model diagnoses the joint probability distribution function (PDF) of these properties via their higher-order statistical moments. Moreover, this framework allows for a very natural PDF-based detrainment closure. While elegant, this modeling approach is complicated and numerical/computational challenges will be discussed. Results from large eddy simulation-based Lagrangian particle tracking experiments will be presented to help motivate and validate this approach.
May 10, 2016
Atmospheric aerosols from natural and anthropogenic sources contribute directly to the Earth’s radiative balance. Additionally, aerosols can also indirectly impact the radiative balance through interactions with clouds. Recently, policies have been enacted to improve air quality and reduce air pollutants for human health concerns. The anthropogenic contribution to atmospheric aerosol loading is presumed altered as a by-product in regions with such policies.
In this work we examine trends in aerosol loading over the western North Atlantic Ocean from 2000 to 2012. This period is notable due to emissions control measures in place in the U.S. Using MODIS satellite observations, we find a -0.02 to -0.04 per decade trend in aerosol optical depth (AOD) in the mid-latitudes. This is further identified as anthropogenic in origin using GOCART model simulations and IMPROVE ground observations. Aerosol-cloud interactions are explored given this AOD decrease and cotemporaneous cloud effective radius increase. We show a domain-wide agreement with Twomey theory, yet reveal “counter-Twomey” behavior in the near costal region and explore reasons as to why. Our results suggest the decreasing aerosol load leads to a decrease in outgoing radiation at the top of the atmosphere over the western North Atlantic region.
May 10, 2016
Commercial marine vessels (CMVs) emit significant amounts of NOX, an ozone precursor, which can contribute to health problems for people living near shipping lanes. In coastal US states, many metropolitan areas such as Baltimore and New York City are located near ports with CMVs. In order to model air quality, the EPA develops emissions inventories for a variety of emission sectors such as on-road vehicles, power plants, and CMVs. In this study, we focus our analysis on the CMV emissions sector, making necessary adjustments to better represent these emissions within the model framework. CMAQ simulations with the adjusted CMV emissions are executed for summer months in 2011 and compared to baseline simulations. Emissions are also projected forward for 2018, and the relative contribution of these adjustments on surface air quality is examined. The long-term goal is to develop top-down emissions inventories from satellite data to compare to the EPA developed emissions inventories.