1,720,996 research outputs found
Dynamic and Thermodynamic Controls on the Amount and Distribution of Orographic Precipitation
No full textThesis (Ph.D.)--University of Washington, 2015This thesis examines various mechanisms controlling the amount and distribution of precipitation in mountainous terrain. In Chapter 2, linear theory and numerical simulations are used to investigate how the tropopause affects the vertical structure of mountain waves, and in turn orographic precipitation. In idealized numerical simulations, variations in the height of the tropopause are found to strongly modulate the depth and magnitude of windward ascent, resulting in a factor-of-two difference in total precipitation across a ridge. The implications of this result are then extended to realistic terrain using a modified version of the Smith-Barstad orographic precipitation model. Chapter 3 addresses the causes of rain-shadow variability in the Washington Cascades. Fluctuations in the large-scale circulation over the North Pacific is found to explain around 70\% of interannual variability in the wintertime rain shadow. Across individual storms, the strongest rain shadows are found to be associated with warm-sector events, while the weakest rain shadows occur during warm-frontal passages. Inter-storm variability in the Cascade rain shadow is further explored in Chapter 4, via numerical simulations involving both real and idealized terrain. Storms with weak rain shadows are shown to exhibit much less evaporation east of the crest than storms with strong rain shadows. The suppression of east-slope evaporation during weak-rain-shadow storms is found to be caused by the presence of stagnant, stable air at low levels, which in turn is shown to be a consequence of warm-frontal passage. In Chapter 5, a simple numerical model is used to evaluate the response of orographic precipitation to surface warming under idealized conditions representative of some of the strongest orographic storms. An upward shift is found in the pattern of condensation with warming, caused by larger fractional changes in condensation at low temperature and amplified warming aloft. As a result, the distribution of precipitation shifts downwind, causing larger fractional changes in precipitation on the lee slope than on the windward slope. In addition, total precipitation is found to increase by a smaller fraction than near-surface water vapor, in contrast to expected changes in other types of extreme precipitation
A Paradigm Shift in Precipitation Modeling: Moving Beyond Numerical Models
No full textThesis (Master's)--University of Washington, 2025Accurately representing surface precipitation in weather and climate modeling is crucial to practical operational use of these models. Presently, global numerical weather prediction (NWP) models struggle to recreate the precipitation variable due to unresolved physical processes at the subgrid level as well as use of poorly constrained microphysical parameterizations. The advent of machine learning in the field of weather and climate prediction has proven to be beneficial to advance modeling efforts and has the potential to also target weak points in NWP such as estimating the precipitation field. Training a deep learning model using satellite data, we can bypass the parameterizations traditionally used by NWP to produce precipitation, achieving a field that more closely matches observations than the widely used ERA5 reanalysis dataset. The resulting model can compute precipitation from only ten ERA5 input fields and is able to better capture extremes while also improving the issue of overproduction of light precipitation in the ERA5 product when evaluated against the IMERG satellite dataset. The machine learning model is also used to produce precipitation from NWP forecast fields, improving on the forecasted precipitation of the NWP model. This work supports future development of deep learning models that meet the needs of current weather and climate modeling
The Influence of an Orographic Feature on an Idealized Mid-Latitude Cyclone
No full textThesis (Ph.D.)--University of Washington, 2017-08The interaction of a mid-latitude cyclone with an isolated north-south mountain barrier is examined using idealized numerical simulation. A prototypical cyclone develops from an isolated disturbance in a baroclinically unstable shear flow upstream of the ridge, producing a cold front that interacts strongly with the topography. The structure and evolution of the lee waves launched by the topography are analyzed, including their temporal and their north-south variation along the ridge. Typical mountain wave patterns are generated by a 500-m high mountain, while substantial wave breaking occurs above a mountain with 2-km height, both at low levels in the lee and in the lower stratosphere. Both local wave characteristics (like their structure and magnitude) and integrated effects of these waves (the pressure drag and momentum flux) often exhibit significant differences from the waves produced in 2D or 3D simulations with steady large-scale flow structures corresponding to the instantaneous conditions over the mountain in the evolving flow. Stratospheric wave breaking over a sufficiently high ridge causes significant removal of the cross-mountain momentum, and strong regions of deceleration are observed above the jet core. Low-level blocking and displacement of the developing cyclone are other ways the mountain influences the synoptic system. Small-amplitude perturbations strongly influence the domain-averaged flow response to the terrain. The mountain waves are also observed to have an influence on the atmospheric KE spectra, producing a k^{-5/3} spectra over a wide range of mesoscale wavenumbers in the stratosphere; this slope is not present in the absence of terrain. The spectral energy budget is calculated, and the gravity waves directly inject energy into the mesoscale, which is then cascaded upscale. These results suggest that terrain forcing is sufficient for building a k^{-5/3} slope, and that direct forcing of the mesoscale is necessary for production of the observed mesoscale slopes, invalidating inertial cascade assumptions
Higher Order Trapped Lee Waves in the Stratosphere
No full textThesis (Master's)--University of Washington, 2018Wave activity with maximum amplitude in the stratosphere downstream of major mountain ranges has previously been explained as a result of wave breaking inducing secondary wave generation in the stratosphere (Vadas et al., 2003) or the vertical propagation of low-amplitude waves from below. However, theoretical results as early as Scorer (1949) and Corby and Wallington (1956) have posited the existence of higher-order trapped wave modes. Although the higher-order modes in these studies were confined to the troposphere due to the studies’ theoretical setup, Corby and Wallington (1956) demonstrate that the level of maximum amplitude for higher-order modes in a two-layer atmosphere with constant Scorer parameter in each layer is higher in the vertical than that of lower-order modes. The increased height of the level of maximum amplitude raises the possibility of higher-order modes with significant amplitude in the stratosphere. However, while higher-order trapped wave modes are theoretically predicted to exist, they are not known to have been observed in the real atmosphere or simulated in realistic numerical models. The DEEPly Propagating Gravity WAVE campaign (DEEPWAVE) was conducted over New Zealand from 29 May 2014 to 27 July 2014 (Fritts et al., 2015). Immediately after the end of the intensive observing period, on 28 July, a strong event occurred. Model simulations of this event revealed unusual wave activity in the stratosphere. These waves were located downstream of the topography, like a trapped wave, but were oriented from south to north, in contrast to the more typical southwest-northeast orientation paralleling the crest of the Southern Alps. Vertical cross-sections of the unusual waves exhibit a nodal structure consistent with that of a higher-order trapped wave mode. Solutions to the two-dimensional, linear, Boussinesq wave equation for a horizontally homogeneous sounding derived from the 28 July case in- clude higher-order modes supported by the zonal wind which have large amplitude in the stratosphere. These higher-order modes are trapped by very strong westerly winds in the upper stratosphere. In contrast, the cross-mountain wind component is not strong enough in the stratosphere to trap the same wave mode in a crest-parallel orientation. These waves are reproducible in both two- and three-dimensional compressible numerical models with both idealized and realistic terrains, and therefore provide a plausible explanation for the wave activity in the stratosphere
A Sea Surface Model for Coupled Data-Driven S2S Forecasting
No full textThesis (Master's)--University of Washington, 2023Data-driven modelling of the atmosphere has rapidly become a vibrant area of research. Recent studies have shown these models have the ability to outperform existing state-of-the-artnumerical weather prediction models. Many of these efforts, however, remain targeted at
short range forecasts (within 2 weeks). We propose using recent advancements in machine
learning to extend the window of predictive skill to the seasonal to subseasonal timescales
(2-10 weeks). To do this we believe capturing couplings between Earth system components
is necessary. Toward this end we have developed an entirely data-driven sea surface model.
Our model predicts global sea surface temperature at daily resolution and can be run iterative like traditional circulation models. We find that even without atmospheric influence,
our ocean model can produce skillful forecasts, consistently beating persistence and outperforming a climatological forecast out to 60 days. We also succeed in predicting the extreme
El Nino event of 2015 at extended leadtimes. Interestingly, our models can run freely for over
a year without producing unstable behavior even though they have no prescribed physical
constraints such as conservation of energy. Furthermore we show that adding information
about the atmosphere can significantly improve upon model performance suggesting that
a these architectures are capable of learning coupled atmosphere-ocean interactions. Our
study is an important step toward developing a fully coupled Deep Learning Earth System
Model
Orographic Precipitation Enhancement in the Presence of Large-Scale Weather Systems
No full textThesis (Ph.D.)--University of Washington, 2022Heavy precipitation in midlatitude mountain ranges is driven by the episodic passage of weather systems. It is a common paradigm that this precipitation results from the large inte- grated vapor transport (IVT) observed in the warm-sectors of these storms. Consequentially, prior idealized simulations of midlatitude orographic precipitation have simulated precipita- tion events with unidirectional, vertically sheared airstreams encountering terrain. Here we refute the hypothesis that the dominant control on orographic precipitation is the upstream IVT. We compare the IVT and resulting precipitation in published observations of real-world events with the same quantities in published studies of idealized moist airstreams impinging on a ridge. We find that numerical simulations of the upslope precipitation produced by moist horizontal unidirectional environmental winds, including 2D, 3D and real topography, tend to significantly underestimate the precipitation in comparison to actual events having the same IVT, suggesting an alternative driver of precipitation.To explore the dynamic mechanisms behind this orographic precipitation enhancement during storm events, we conducted a novel set of idealized numerical simulations. This set of simulations was designed to isolate the effects of mesoscale circulations induced by the terrain modified flow from the effect of forced uplift of a moist airstream alone.
We conducted an idealized simulation of an archetypal midlatitude cyclone encountering an isolated 1-km tall, 500-km long, North-South topographic ridge. The cyclone is initialized with a PV anomaly which develops into a mature cyclone over the course of 2.5 days. Our focus is on the enhancement of the orographic precipitation as the warm sector of the storm traverses the ridge. The unique idealized setup of the simulation allows for two addi- tional, closely related experiments to be conducted. In the first additional experiment, the cyclone simulation is run with flat terrain. In the second, a unidirectional, parallel shear flow simulation is run with the flow encountering the same idealized ridge as in the original experiment. The shear flow is initialized with a vertical profile of wind, temperature, and moisture taken from the warm sector of the midlatitude cyclone, and therefore has the same cross-terrain integrated vapor transport (IVT) as the Cyc+Mt case. Hereafter, the simula- tion with both the cyclone and the ridge, the flat terrain cyclone simulation, and the parallel shear flow simulations will be referred to as the Cyc+Mt, Cyc-Flat, and Shear simulations, respectively.
We found that there was significantly more precipitation on the upslope side of the terrain in the case with the cyclone and the mountain compared to the shear flow case. This precipitation enhancement is strongly correlated in space and time with ridge-scale low-level moisture flux convergence (MFC) and the enhancement of embedded convection in the nearly neutrally stratified moist airstream. The large MFC in the Cyc+Mt simulation stems from two features of the mesoscale flow pattern produced by the interaction of the large-scale disturbance with the terrain. The first arises from differences in the mountain- wave structure in comparison to the shear-flow case. The cross-mountain flow at the crest is slower than in the shear flow case, thereby increasing the East-West mass-and moisture- flux convergence over the windward slope. Second, a low-level jet on the upslope side of the terrain in the Cyc+Mt case brings in warm moist air that increases the North-South moisture flux convergence on the windward slope relative to the shear flow case.
The net result of these differences in moisture-flux convergence is that there is about four times more precipitation over the mountain in the warm sector of the cyclone and mountain simulation than in the shear flow simulation, despite the two simulations having identical upstream IVT. The precipitation generated by the cyclone in the absence of the mountain is also far weaker, with nearly zero precipitation accumulation in the warm sector of the storm until the arrival of the front. Our simulations demonstrate that high values of upstream IVT are not responsible for large orographic precipitation values, and that mesoscale features of the terrain-induced flow control rainfall
Mesoscale Predictability and Error Growth in Short Range Ensemble Forecasts
No full textThesis (Master's)--University of Washington, 2013Although it was originally suggested that small-scale, unresolved errors corrupt forecasts at all scales through an inverse error cascade, some authors have proposed that those mesoscale circulations resulting from stationary forcing on the larger scale may inherit the predictability of the large-scale motions. Further, the relative contributions of large- and small-scale uncertainties in producing error growth in the mesoscales remain largely unknown. Here, 100 member ensemble forecasts are initialized from an ensemble Kalman filter (EnKF) to simulate two winter storms impacting the East Coast of the United States in 2010. Four verification metrics are considered: the local snow water equivalence, total liquid water, and 850 hPa temperatures representing mesoscale features; and the sea level pressure field representing a synoptic feature. It is found that while the predictability of the mesoscale features can be tied to the synoptic forecast, significant uncertainty existed on the synoptic scale at lead times as short as 18 hours. Therefore, mesoscale details remained uncertain in both storms due to uncertainties at the large scale. Additionally, the ensemble perturbation kinetic energy did not show an appreciable upscale propagation of error for either case. Instead, the initial condition perturbations from the cycling EnKF were maximized at large scales and immediately amplified at all scales without requiring initial upscale propagation. This suggests that relatively small errors in the synoptic-scale initialization may have more importance in limiting predictability than errors in the unresolved, small-scale initial conditions
Atmospheric predictability is insensitive to the slope of the background kinetic energy spectrum
No full textThesis (Master's)--University of Washington, 2021The sensitivity of atmospheric predictability to the slope of the background kinetic energy (KE) spectrum is investigated by adding low-level potential temperature perturbations of varying scales and amplitudes to convection-permitting idealized simulations of moist baroclinic waves using the Weather Research and Forecasting (WRF) model. When perturbations are small in amplitude, error growth through 36 hour lead times is insensitive to initial error scale and the slope of the background KE spectrum. In physical space, this insensitivity occurs because short-range predictability is limited by rapid growth from moist processes such as cold frontal convection, regardless of the initial error scale. In spectral space, this insensitivity occurs because errors grow up-amplitude at all scales rather than via an explicit upscale cascade from the smallest scales. These results are confirmed using two different baroclinic wave simulations: one with a widely-used but unrealistic base state which pro- duces a flat KE spectrum in the mesoscales, and another with more realistic moist processes and a k^{−2} mesoscale KE spectrum. The observed insensitivity of error growth in moist baroclinic waves to the slope of the background KE spectrum contrasts with error growth in homogeneous isotropic turbulence, which is local in spectral space and highly sensitive to the KE spectral slope
The downstream decay of trapped lee waves
No full textThesis (Ph.D.)--University of Washington, 2014The mechanisms through which trapped lee waves decay, and where this decay occurs, are of utmost importance in order to understand the impact that these waves have on the larger-scale climate system. Previous studies have shown trapped waves as contributing a significant fraction of the total orographic drag, but they remain poorly understood. In this dissertation, two decay mechanisms are analyzed and compared --- stratospheric leakage, and boundary layer absorption. Decay of lee waves through upward leakage of wave energy towards the stable stratosphere is studied primarily using a linear Boussinesq model, forced by either a three-layer atmosphere or a more realistic four-layer atmosphere containing vertical wind shear and an elevated inversion. Weak downstream decay occurs due to the stratosphere in the highly-idealized three-layer atmosphere, albeit at too slow of a rate for the typical decay seen in nature. In the more realistic profile, rapid downstream decay occurs through stratospheric leakage --- leading to a removal of the wavetrain within 1.5 wavelengths in the most extreme case of a 200 m deep elevated inversion. As the depth the elevated inversion is reduced, the potential rate of downstream decay is increased. For all profiles, the rate of leakage due to the stratosphere is shown to be maximized for values of stratospheric stability (N<sub>s</sub>) slightly larger than for the threshold for decay, with a decreasing trend in the rate of decay as the stratospheric stability is further increased. The impact of the stratosphere and boundary layer on trapped wave decay are both simulated using a full nonlinear numerical model. Decay through boundary layer absorption is seen to vary slightly with the atmospheric profile --- relating to the location and the structure of the resonant wave duct compared to the boundary layer. Rates of downstream decay due to the stratosphere agree well between the linear and nonlinear models. Given the highly-idealized atmospheric profile, boundary layer decay is dominant with minimal decay occurring through stratospheric leakage at any N<sub>s</sub>. With the realistic profile shown by the linear model to be suitable for strong stratospheric leakage, downstream decay is stronger due to the stratosphere than for the roughest lower boundary simulated (z<sub>0</sub> = 0.5 m, where z<sub>0</sub> is the roughness length). A move towards understanding the decay of trapped waves in three dimensions is also discussed through use of high-resolution simulations of lee waves downwind of the Aleutian islands using WRF. In the control run, close agreement is found between the modeled wave field, and that observed by satellite. As the roughness length of the lower boundary is increased, the rate of decay is noted to increase by approximately 10% across the range of z<sub>0</sub> simulated --- although much of this increase occurs across the change from 10<\super>-2</super> m to 10<super>-1</super> m, rather than the more linear increase seen in our 2D simulations. An additional subject discussed is the generation of striations in stacked lenticular clouds. High-resolution numerical simulations show that striations in excess of 150 m in width may be generated by perturbations in the relative humidity as small as ± 0.25%. Perturbations of this scale are small enough to be likely ubiquitous in nature, explaining why these clouds always have a layered appearance
Orographic Precipitation in an Idealized Midlatitude Cyclone
No full textThesis (Master's)--University of Washington, 2019In mountainous areas in the midlatitudes, the majority of terrain related precipitation occurs in conjunction with large scale storm systems. As noted in Smith 2006, ``Mountains have their most profound influence on precipitation during brief events when significant atmospheric disturbances move into mountainous areas''. Despite the established importance of synoptic dynamics, idealized studies in the past have generally neglected the large scale flow. The difficulty of modeling a complex synoptic scale system in an idealized numerical simulation has been a barrier. In this study, we aim to rectify this by presenting a series of idealized experiments of a realistic, prototypical midlatitude cyclone encountering an idealized ridge. The simulation is configured to have a canonical midlatitude cyclone located in an east-west channel flow. The cyclone encounters an isolated 2-km ridge after 2.5 days of simulation time. Two different ridge orientations were used in different regions of the cyclone; a north-south barrier was placed parallel to the front of the cyclone, and in a separate experiment, an east-west barrier was placed in the warm sector of the cyclone. In addition to the cyclone-and-mountain simulations, two control test cases were conducted. One test case included the cyclone but had flat terrain; the other test case did not include the cyclone, but was initialized with parallel shear flow over the terrain. For the simulation with the north-south mountain and the cyclone, simulations conducted with microphysics show an enhancement of precipitation over the mountain equal to the sum of the flat terrain and the shear-flow-only test cases. For the east-west mountain case, the precipitation observed in the cyclone+mountain case was 1.5-2 times the total combined precipitation in the flat-terrain and shear flow cases. In the case of the east-west ridge, the precipitation enhancement resulted from the updrafts due to the flow over terrain interacting with vertical motion in the larger system. The seeder-feeder effect also played a role. In the case of the north-south ridge along the front, the broadening and deepening of convective updrafts contributed to an increase in precipitation. In both cases, horizontal velocity gradients present in the cyclone generated greater convergence and up slope flow than just forced shear flow control experiments
- …
