Improved observations of the atmospheric boundary layer (noted as ABL or BL herein) and its interactions with the ocean, land, and ice surfaces have great potential to advance science on a number of fronts, from improving forecasts of severe storms and air quality to constraining estimates of trace gas emissions and transport. Understanding the BL is a crucial component of model advancements, and increased societal demands for extended weather impact forecasts (from hours to months and beyond) highlight the need to advance Earth system modeling and prediction. New observing technologies and approaches (including in situ and ground-based, airborne, and satellite remote sensing) have the potential to radically increase the density of observations and allow new types of variables to be measured within the BL, which will have broad scientific and societal benefits.
In October 2017, the Board on Atmospheric Sciences and Climate together with the Ocean Studies Board and the Board on Life Sciences of the National Academies of Sciences, Engineering, and Medicine (the National Academies) convened a workshop1 to explore the future of BL observations and their role in improving modeling and forecasting capabilities. Workshop participants discussed the science and applications drivers for BL observation, emerging technology to improve observation capabilities, and strategies for the future. This document summarizes workshop presentations and plenary discussions.
The atmospheric boundary layer is the lowest kilometer or so of the atmosphere closest to Earth’s surface. It has substantial variability due to, for example, the nature of the underlying surface, diurnal cycling, thermal stratification, vertical entrainment, and advective processes (see Figure 1). This portion of the atmosphere interacts with Earth’s land, ocean, and ice surfaces, and these interactions control exchanges of kinetic and thermal energies and materials such as aerosols and pollutants at the surface, which in turn drive the global weather and climate system as a whole. The conditions in the ABL can influence, for example, how much rain will fall during a storm, where and how far pollutants will spread, or the strength and path of hurricanes and tornadoes.
Accurate measurements and better understanding of BL processes are of key importance for advancing many areas of science and a host of applications. Most directly, advancements could be made to improve weather and climate models and predictions. Increased societal demands for extended weather forecasts (from hours to
1 This Proceedings of a Workshop was prepared by the workshop rapporteur as a factual summary of what occurred at the workshop. The planning committee’s role was limited to planning and convening the workshop. See the appendixes for the Statement of Task, planning committee biosketches, workshop agenda, and participant list.
months and beyond) make this an ideal time to examine new understanding of BL processes as well as new technologies for measurements. In addition, many of the same sorts of observations that are expected to advance weather and climate prediction would also enable improvements in the ability to track and predict the distribution of trace gases, aerosols, and other particles in the atmosphere. Improved atmospheric measurements could also enhance understanding of the exchanges between the biosphere and the atmosphere, and likewise the air-sea exchanges, which are critical to improved understanding of the climate system. Better understanding these exchange processes is important for scientific knowledge of biogeochemical cycles, impacts of climate change on ecological systems, air pollution, and estimates of carbon storage in natural systems, among many other applications.2,3
Understanding of the BL, and in turn the ability to make significant advances in many application areas, is constrained by the lack of observations at sufficient spatial coverage and temporal resolution. Global satellite observations are key for sampling boundary conditions over oceans and providing information for global weather predictions; however, their ability to penetrate and resolve the vertical structure in the BL is still limited. Technology for ground-based active and passive profilers exists to provide adequate coverage of the evolution of BL height, winds, temperature, and humidity over land down to the mesoscale, as recommended by previous National
2 National Academies of Sciences, Engineering, and Medicine. 2016. The Future of Atmospheric Chemistry Research: Remembering Yesterday, Understanding Today, Anticipating Tomorrow. Washington, DC: The National Academies Press. doi: https://doi.org/10.17226/23573.
3 National Research Council. 2010. Verifying Greenhouse Gas Emissions: Methods to Support International Climate Agreements. Washington, DC: The National Academies Press. doi: https://doi.org/10.17226/12883.
Academies studies.4,5 These technologies can help enable understanding of BL processes and improve weather and air quality prediction. However, this capability has yet to be widely used and is rarely deployed over the ocean or polar regions.
A number of emerging technologies provide opportunities for advancement in BL observing. One key focus of the workshop involved applying these emergent technological capabilities to relevant science questions and applications and to leverage existing networks to address open BL observation needs. To make progress in these areas, the planning committee challenged the participants to consider the following overarching science question during the workshop: What observations are needed (over land, ocean, and ice) to make meaningful progress in our understanding and modeling of the global atmospheric boundary layer? Following a series of keynote remarks intended to provide an overview of research to date, a number of focused panel discussions were held to review the science and applications drivers for BL research and emerging technologies for observations. Panels considered a wide range of topics including weather, subseasonal to seasonal (S2S), and climate modeling; exchanges between the biosphere and atmosphere; terrestrial and marine BL science questions; modeling and parameterization; and pollutant and greenhouse gas (GHG) emissions, air quality, and human health. Panelists also explored technology options for advancing understanding including optics, photonics, and sensors; in situ measurements; existing networks; and surface- and space-based remote sensing. Detailed summaries of workshop presentations can be found in the speaker abstracts (available at https://ww.nap.edu/catalog/25138 under the Resources tab). Participants also engaged in breakout group discussions on a number of cross-cutting topics such as field programs to improve model physics; integrating observations across platforms; utilizing partnerships; example targeted regions for intensive study; and observations for improved predictions.
Over time, scientific understanding of the structure and definition of the BL has evolved. Keynote speaker Peggy LeMone opened the workshop by providing a historical perspective on BL observations and noted how understanding the history can help guide future observations and research. Although improved measurement and modeling techniques over the past few decades have helped scientists reach new levels of understanding, a number of challenges remain. Examples of these challenges include understanding the nocturnal BL, balancing the surface energy budget, failures of
4 National Research Council. 2009. Observing Weather and Climate from the Ground Up: A Nationwide Network of Networks. Washington, DC: The National Academies Press. doi: https://doi.org/10.17226/12540.
5 National Academies of Sciences, Engineering, and Medicine. 2016. Next Generation Earth System Prediction: Strategies for Subseasonal to Seasonal Forecasts. Washington, DC: The National Academies Press. doi: https://doi.org/10.17226/21873.
the Monin-Obukhov (MO) similarity theory,6 and understanding surface-atmosphere interactions. She pointed out that idealized cases and models do not always represent reality, and thus there is a need for greater observational capabilities. A combination of techniques, including improved observations that allow for different kinds of modeling (such as Large Eddy Simulation [LES] and Mesoscale Models), will be important for future progress and understanding.
Over land, variations in the BL are driven by the diurnal cycle of short wave and long wave radiative processes. To illustrate these processes, Alan Betts described a noteworthy Prairie dataset that demonstrates some key features of the Northern latitude BL and surface climate, where cloud forcing is the dominant BL driver. Cloud radiative forcing changes from negative to positive with snow cover, and snow cover is a fast climate switch between cloud-coupled unstable and stable BLs with distinct non-overlapping climates (see Figure 2). This suggests the importance of the fraction of days with snow cover for the cold season temperature. Seasonal shifts with snow are also important for agricultural models, as they use seasonal forecasts and reanalysis.
When comparing the representation of the Canadian Prairies in the European Centre for Medium-Range Weather Forecasts (ECMWF) model with measurements, Dr. Betts found that model biases in stable and unstable BLs vary with season and cloud cover, snow transitions give step changes in model biases, and the balance between ground coupling and turbulent transports for stable BLs is poorly modeled because it is poorly known. There are few routine measurements of the surface fluxes and BL structure, and satellite observations poorly sample the near-surface BL. Parameterizations should be used to represent the coupled BL processes because they cannot be explicitly resolved except using direct numerical simulation (DNS) or LES, Dr. Betts stated. Many of these were developed for homogeneous conditions (e.g., MO similarity theory for the profiles), and their applicability to real-world heterogeneous conditions are not well tested.
The marine boundary layer (MBL) is characterized by variations in turbulence, radiation, clouds, and interactions between the atmosphere and sea surface. Chris Fairall noted that the marine environment is observationally sparse, requiring that models be employed in coordination with existing observations in studying the BL. For example, wave models can help compute energy flux to the surface through simplified representations. He noted a number of additional reasons to use a model such as difficulty observing specific details; wind-wave parameters that are highly correlated and difficult to separate in observations; extrapolation to higher wind speeds; using models and physics to relate fluxes, waves, and wave breaking properties; and relatively small amounts of data needed (i.e., thousands of hours of data are not required). Observations of fluxes from sources such as buoys and ships are in relative agreement, but the modeling of these fluxes has significant disagreements. Calculation of radiative cloud forcing at the surface using the same data also shows that the models are inaccurate, likely due to incorrect cloud fraction.
6 A relationship describing the vertical behavior of nondimensionalized mean flow and turbulence properties within the atmospheric surface layer (the lowest 10% or so of the atmospheric boundary layer) as a function of the Monin-Obukhov key parameters. See http://glossary.ametsoc.org/wiki/Moninobukhov_similarity_theory.
Dr. Fairall noted that wave breaking is important for gas transfer, sea spray aerosol generation, and hurricane sea spray effects. White cap fraction is extremely important but hard to observe, and turbulent fluxes of heat and moisture in environments with sea ice also add complexity. Bulk variables over water and ice need to be specified, as well as ice properties (fraction of ice cover, freeboard, roughness) and second order effects such as flow size distribution, flow spacing, partially frozen leads, and melt ponds. Problems associated with cloud observations include turbulence in clouds and its ability to drive the entire BL; however, remote sensing is one approach to studying this problem. In flux parameterizations, there are efforts to add wave or ice information to improve the parameterizations. A revolution in observation techniques is needed to obtain better area averaging and ways to process ice information. The wave information gives a better extrapolation mode than using only wind speed. In the near-surface profiles, some puzzling early results are being examined, particularly in the wavy BL or outside of the MO similarity layer. New observing methods such as Unmanned Aerial Systems (UASs, also known as “drones” or “UAVs”) show promise in this area. In cloud-turbulence interactions, there is tremendous potential in multi-wavelength Doppler radar and lidar (either surface-based or on aircraft) with applications in stratocumulus and possibly shallow convection.
Effective models and parameterizations of the ABL go hand in hand with enhanced ABL observations. Chris Bretherton expounded on some of the current opportunities and challenges associated with BL modeling and the linkages to improved observations. Recent development of high-resolution models enables even fairly large-scale models to resolve aspects of circulation that are relevant to the ABL (see Figure 3). Improved topography is also available in high-resolution models and one can see the effects of terrain, for example, on temperature. A high-resolution model is also needed to represent the variability in the BL structure in a region of complex topography, for example, in coastal regions and hilly or mountainous terrain. The finest resolution
models used in process studies are LES models, and these are limited-area models that have a three-dimensional grid with fine enough resolution to explicitly simulate the large eddies that contribute to most vertical turbulent transport. This type of model is a workhorse for ABL process modeling and a reference for ABL parameterization development. Again, observations are crucial in evaluating LES skill in different ABL regimes.
Dr. Bretherton noted that high-resolution models partner well with ABL observations. He discussed high-resolution models that resolve terrain and boundary heterogeneity and as a result, there is less representativeness bias when comparing LES with point observations. There is better vertical resolution, which improves ABL representations and better representation of deep cumulus convection and precipitation. He also pointed to the importance of observations, but unfortunately, better observations do not necessarily guarantee good performance from the model. There are still parameterization, initialization, and numerical challenges, even with high-resolution models. Observations also play a crucial role in evaluation and data assimilation. In global models, ABL metrics focus on the surface (land temperature and
surface radiation, as well as ocean surface wind stress and latent heat flux). Metrics like BL height are usually too inexactly defined to be useful for comparisons of models with observations. Given the importance of ABL vertical stratification, shear, and turbulence structure to air pollution, wind power, aviation, and other applications as well as current surface-based remote profiling instrumentation, Dr. Bretherton suggested that perhaps robust metrics of ABL vertical structure for model diagnostics packages could be developed. Many ABL observations can be assimilated into forecast models. For example, skin temperature (or the temperature of Earth’s land, ocean, and ice surfaces) measured from satellites can be compared with the models, and assimilating the data could potentially improve BL representation. However, over land, the skin temperature is more correlated with the surface type than with air temperature and has little persistence due to the low surface heat capacity especially in the dry sunny regions where it can be retrieved. The skin temperature of the ocean has longer persistence due to its larger heat capacity, which aids in the removal of clouds from infrared (IR) measurements of skin temperature using repeated satellite passages. Over both surfaces, researchers are using passive microwave measurements to estimate the surface temperature, albeit with generally lower resolution than IR estimates. A systematic Observing System Simulation Experiment7 (OSSE)-informed design of an ABL observing system could be one way to approach this issue.
Dr. Bretherton noted that new observations are revealing and important, but often expensive. Collaborating with the modeling community can help maximize investments in better ABL representation in all types of U.S. models. High-resolution models and LES can be good partners for ABL observations, though additional, routine ABL vertical profile observations and model evaluation metrics are needed. Observations could be directed to address specific quantified model weaknesses, and some enduring modeling challenges include clouds, stable BLs, and complex land surfaces.
Some workshop participants noted that, on a small scale, models compare well to field observations. However, when considering regional or global models, the bias is particularly large. Therefore, the scale used is a crucial consideration because the fully coupled system is complex. Global and regional models handle advective processes relatively well, but the BL coupling and the way in which it is parameterized within the regional or global model gives rise to the biases noted. Improving the vertical coupling and its parameterization, whether on the regional scale or on the global scale, remains a challenge. Some participants also noted that the challenge with finer scale models, where observational data exist, is that it is hard to couple them to the information that comes in from the BL. When initializations cannot be measured to the time and space scales of interest, larger scale models are used.
7 An experiment in which a model is assumed to represent the exact behavior of the physical system (the atmosphere) of interest. The model results are then sampled in a fashion to mimic some real or hypothetical observing system. OSSEs can be a way to guide the design of new or enhanced observing systems or to evaluate the usefulness of existing observations. See http://glossary.ametsoc.org/wiki/Observing_systems_simulation_ experiment.
This page intentionally left blank.