Search for: All records

When testing hypotheses about which of two competing models is better, say A and B, the difference is often not significant. An alternative, complementary approach, is to measure how often model A is better than model B regardless of how slight or large the difference. The hypothesis concerns whether or not the percentage of time that model A is better than model B is larger than 50%. One generalized test statistic that can be used is the power-divergence test, which encompasses many familiar goodness-of-fit test statistics, such as the loglikelihood-ratio and PearsonX²tests. Theoretical results justify using thedistribution for the entire family of test statistics, wherekis the number of categories. However, these results assume that the underlying data are independent and identically distributed, which is often violated. Empirical results demonstrate that the reduction to two categories (i.e., model A is better than model B versus model B is better than A) results in a test that is reasonably robust to even severe departures from temporal independence, as well as contemporaneous correlation. The test is demonstrated on two different example verification sets: 6-h forecasts of eddy dissipation rate (m^2/3s⁻¹) from two versions of the Graphical Turbulence Guidance model and for 12-h forecasts of 2-m temperature (°C) and 10-m wind speed (m s⁻¹) from two versions of the High-Resolution Rapid Refresh model. The novelty of this paper is in demonstrating the utility of the power-divergence statistic in the face of temporally dependent data, as well as the emphasis on testing for the “frequency-of-better” alongside more traditional measures.

 

Radiative transfer (RT) is a crucial but computationally expensive process in numerical weather/climate prediction. We develop neural networks (NN) to emulate a common RT parameterization called the Rapid Radiative Transfer Model (RRTM), with the goal of creating a faster parameterization for the Global Forecast System (GFS) v16. In previous work we emulated a highly simplified version of the shortwave RRTM only—excluding many predictor variables, driven by Rapid Refresh forecasts interpolated to a consistent height grid, using only 30 sites in the Northern Hemisphere. In this work we emulate the full shortwave and longwave RRTM—with all predictor variables, driven by GFSv16 forecasts on the native pressure–sigma grid, using data from around the globe. We experiment with NNs of widely varying complexity, including the U-net++ and U-net3+ architectures and deeply supervised training, designed to ensure realistic and accurate structure in gridded predictions. We evaluate the optimal shortwave NN and optimal longwave NN in great detail—as a function of geographic location, cloud regime, and other weather types. Both NNs produce extremely reliable heating rates and fluxes. The shortwave NN has an overall RMSE/MAE/bias of 0.14/0.08/−0.002 K day⁻¹for heating rate and 6.3/4.3/−0.1 W m⁻²for net flux. Analogous numbers for the longwave NN are 0.22/0.12/−0.0006 K day⁻¹and 1.07/0.76/+0.01 W m⁻². Both NNs perform well in nearly all situations, and the shortwave (longwave) NN is 7510 (90) times faster than the RRTM. Both will soon be tested online in the GFSv16.

Radiative transfer is an important process for weather and climate. Accurate radiative transfer models exist, such as the RRTM, but these models are computationally slow. We develop neural networks (NNs), a type of machine learning model that is often computationally fast after training, to mimic the RRTM. We wish to accelerate the RRTM by orders of magnitude without sacrificing much accuracy. We drive both the NNs and RRTM with data from the GFSv16, an operational weather model, using locations around the globe during all seasons. We show that the NNs are highly accurate and much faster than the RRTM, which suggests that the NNs could be used to solve radiative transfer inside the GFSv16.

 

Abstract. Accurate boundary layer temperature and humidity profiles are crucial for successful forecasting of fog, and accurate retrievals of liquid water path are important for understanding the climatological significance of fog. Passive ground-based remote sensing systems such as microwave radiometers (MWRs) and infrared spectrometers like the Atmospheric Emitted Radiance Interferometer (AERI), which measures spectrally resolved infrared radiation (3.3 to 19.2 µm), can retrieve both thermodynamic profiles and liquid water path. Both instruments are capable of long-term unattended operation and have the potential to support operational forecasting. Here we compare physical retrievals of boundary layer thermodynamic profiles and liquid water path during 12 cases of thin (LWP<40 g m−2) supercooled radiation fog from an MWR and an AERI collocated in central Greenland. We compare both sets of retrievals to in-situ measurements from radiosondes and surface-based temperature and humidity sensors. The retrievals based on AERI observations accurately capture shallow surface-based temperature inversions (0–10 m a.g.l.) with lapse rates of up to −1.2 ∘C m−1, whereas the strength of the surface-based temperature inversions retrieved from MWR observations alone are uncorrelated with in-situ measurements, highlighting the importance of constraining MWR thermodynamic profile retrievals with accurate surface meteorological data. The retrievals based on AERI observations detect fog onset (defined by a threshold in liquid water path) earlier than those based on MWR observations by 25 to 185 min. We propose that, due to the high sensitivity of the AERI instrument to near-surface temperature and small changes in liquid water path, the AERI (or an equivalent infrared spectrometer) could be a useful instrument for improving fog monitoring and nowcasting, particularly for cases of thin radiation fog under otherwise clear skies, which can have important radiative impacts at the surface. 

Abstract Land-atmosphere feedbacks are a critical component of the hydrologic cycle. Vertical profiles of boundary layer temperature and moisture, together with information about the land surface, are used to compute land-atmosphere coupling metrics. Ground based remote sensing platforms, such as the Atmospheric Emitted Radiance Interferometer (AERI), can provide high temporal resolution vertical profiles of temperature and moisture. When co-located with soil moisture, surface flux, and surface meteorological observations, such as at the Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) site, it is possible to observe both the terrestrial and atmospheric legs of land-atmosphere feedbacks. In this study, we compare a commonly used coupling metric computed from radiosonde-based data to that obtained from the AERI to characterize the accuracy and uncertainty in the metric derived from the two distinct platforms. This approach demonstrates the AERI’s utility where radiosonde observations are absent in time and/or space. Radiosonde and AERI based observations of the Convective Triggering Potential and Low-Level Humidity Index (CTP-HI low ) were computed during the 1200 UTC observation time and displayed good agreement during both 2017 and 2019 warm seasons. Radiosonde and AERI derived metrics diagnosed the same atmospheric preconditioning based upon the CTP-HI low framework a majority of the time. When retrieval uncertainty was considered, even greater agreement was found between radiosonde and AERI derived classification. The AERI’s ability to represent this coupling metric well enabled novel exploration of temporal variability within the overnight period in CTP and HI low . Observations of CTP-HI low computed within a few hours of 1200 UTC were essentially equivalent, however with greater differences in time arose greater differences in CTP and HI low . 

Abstract. During the Chequamegon Heterogeneous Ecosystem Energy-balance Study Enabled by a High-density Extensive Array of Detectors 2019 (CHEESEHEAD19) field campaign, held in the summer of 2019 in northern Wisconsin, USA, active and passive ground-based remote sensing instruments were deployed to understand the response of the planetary boundary layer to heterogeneous land surface forcing. These instruments include radar wind profilers, microwave radiometers, atmospheric emitted radiance interferometers, ceilometers, high spectral resolution lidars, Doppler lidars, and collaborative lower-atmospheric mobile profiling systems that combine several of these instruments. In this study, these ground-based remote sensing instruments are used to estimate the height of the daytime planetary boundary layer, and their performance is compared against independent boundary layer depth estimates obtained from radiosondes launched as part of the field campaign. The impact of clouds (in particular boundary layer clouds) on boundary layer depth estimations is also investigated. We found that while all instruments are overall able to provide reasonable boundary layer depth estimates, each of them shows strengths and weaknesses under certain conditions. For example, radar wind profilers perform well during cloud-free conditions, and microwave radiometers and atmospheric emitted radiance interferometers have a very good agreement during all conditions but are limited by the smoothness of the retrieved thermodynamic profiles. The estimates from ceilometers and high spectral resolution lidars can be hindered by the presence of elevated aerosol layers or clouds, and the multi-instrument retrieval from the collaborative lower atmospheric mobile profiling systems can be constricted to a limited height range in low-aerosol conditions. 

Supercooled fogs can have an important radiative impact at the surface of the Greenland Ice Sheet, but they are difficult to detect and our understanding of the factors that control their lifetime and radiative properties is limited by a lack of observations. This study demonstrates that spectrally resolved measurements of downwelling longwave radiation can be used to generate retrievals of fog microphysical properties (phase and particle effective radius) when the fog visible optical depth is greater than ∼0.25. For 12 cases of fog under otherwise clear skies between June and September 2019 at Summit Station in central Greenland, nine cases were mixed‐phase. The mean ice particle (optically‐equivalent sphere) effective radius was 24.0 ± 7.8 µm, and the mean liquid droplet effective radius was 14.0 ± 2.7 µm. These results, combined with measurements of aerosol particle number concentrations, provide evidence supporting the hypotheses that (a) low surface aerosol particle number concentrations can limit fog liquid water path, (b) fog can act to increase near‐surface aerosol particle number concentrations through enhanced mixing, and (c) multiple fog events in quiescent periods gradually deplete near‐surface aerosol particle number concentrations.

 

Abstract. This study presents the first full annual cycle (2019–2020) of ambient surface aerosol particle number concentration measurements (condensationnuclei > 20 nm, N20) collected at Summit Station (Summit), in the centre of the Greenland Ice Sheet (72.58∘ N, −38.45∘ E; 3250 ma.s.l.). The mean surface concentration in 2019 was 129 cm−3, with the 6 h mean ranging between 1 and 1441 cm−3. The highest monthly mean concentrations occurred during the late spring and summer, with the minimum concentrations occurring in February (mean: 18 cm−3). High-N20 events are linked to anomalous anticyclonic circulation over Greenland and the descent of free-tropospheric aerosol down to the surface, whereas low-N20 events are linked to anomalous cyclonic circulation over south-east Greenland that drives upslope flow and enhances precipitation en route to Summit. Fog strongly affects particle number concentrations, on average reducing N20 by 20 % during the first 3 h of fog formation. Extremely-low-N20 events (< 10 cm−3) occur in all seasons, and we suggest that fog, and potentially cloud formation, can be limited by low aerosol particle concentrations over central Greenland. 

Abstract There has been a recent wave of attention given to atmospheric bores in order to understand how they evolve and initiate and maintain convection during the night. This surge is attributable to data collected during the 2015 Plains Elevated Convection at Night (PECAN) field campaign. A salient aspect of the PECAN project is its focus on using multiple observational platforms to better understand convective outflow boundaries that intrude into the stable boundary layer and induce the development of atmospheric bores. The intent of this article is threefold: 1) to educate the reader on current and future foci of bore research, 2) to present how PECAN observations will facilitate aforementioned research, and 3) to stimulate multidisciplinary collaborative efforts across other closely related fields in an effort to push the limitations of prediction of nocturnal convection.