3
National Needs for Mesoscale Observations in Five Economic Sectors

Chapter 2, along with Appendix A, developed a rationale for spacing and frequency of observations that is appropriate for nationwide weather prediction, climate monitoring, and supporting research. This is the most fundamental application of mesoscale observations in that it seraves many other applications of huge economic import, either directly or indirectly.

This chapter exposes the needs for mesoscale observations in five areas vital to the nation’s well-being: (1) energy security, (2) public health and safety, (3) transportation, (4) water resources, and (5) food production. Other areas could have been examined, for example, outdoor recreation or construction, but these five are representative of both the astounding diversity of need and the tremendous value of well-placed and timely observations.

For each sector, we note its importance to the national economy, list current assets and operational requirements for mesoscale observations, and highlight the more critical future needs.

ENERGY SECURITY

Importance to the National Economy

Sufficient and reliable supplies of energy are critical to the security of the nation and for sustained, uninterrupted economic growth. To address the meteorological observations needed to support energy security, we first identify the stages comprising the transition from the primary energy source to final consumption. For example, the use of fossil fuels for energy



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 42
3 National Needs for Mesoscale Observations in Five Economic Sectors Chapter 2, along with Appendix A, developed a rationale for spacing and frequency of observations that is appropriate for nationwide weather prediction, climate monitoring, and supporting research. This is the most fundamental application of mesoscale observations in that it seraves many other applications of huge economic import, either directly or indirectly. This chapter exposes the needs for mesoscale observations in five areas vital to the nation’s well-being: (1) energy security, (2) public health and safety, (3) transportation, (4) water resources, and (5) food production. Other areas could have been examined, for example, outdoor recreation or construction, but these five are representative of both the astounding diversity of need and the tremendous value of well-placed and timely observations. For each sector, we note its importance to the national economy, list current assets and operational requirements for mesoscale observations, and highlight the more critical future needs. ENERGY SECURITY Importance to the National Economy Sufficient and reliable supplies of energy are critical to the security of the nation and for sustained, uninterrupted economic growth. To address the meteorological observations needed to support energy security, we first identify the stages comprising the transition from the primary energy source to final consumption. For example, the use of fossil fuels for energy 

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS consumption involves extraction, refinement, transport, conversion, trans- mission (for electricity), and final end use. Meteorological measurements may or may not be needed for decision making during each of these stages. For instance, the use of coal for generating electricity for residential con- sumption is relatively weather insensitive (except for disruption by extreme events) at the extraction, refinement, and transport stages, but is weather sensitive for conversion (load planning), transmission (routing and line exposure to weather), and final end use (weather-driven demand). Recent trends to replace fossil fuels with renewable primary energy sources—particularly biomass, direct solar, wind, and hydroelectric—create somewhat different transition stages from primary source to final con- sumption and hence somewhat different environmental monitoring needs. Biomass has perhaps the highest vulnerability to weather and therefore the greatest need for weather monitoring, starting with the seasonal weather outlook that favors the planting of one biomass crop over another. Planting, growth, harvest, and transport to consolidation point or conversion facility all present weather-related vulnerabilities not applicable to fossil fuels and call for reliable meteorological measurements and the best-available weather forecasts and seasonal climate outlooks. Direct solar, wind, and hydroelectric power have somewhat different but less complex weather data requirements. In total, renewable primary energy sources call for a wider range of measurements (e.g., soil moisture, direct and diffuse radiation, vertical profiles of wind, snow depth, stream flow, reservoir temperature) at more locations and advances in short-term and seasonal forecasts. In addition, renewable primary energy sources are more vulnerable to extreme events, particularly drought, hail, flood, extreme heat and cold, tornados and hurricanes, than are fossil energy sources. The emerging wind power industry has meteorological observing needs that are similar to other currently unmet needs discussed elsewhere in this report, particularly observations in the lower part of the atmospheric boundary layer above the surface. Wind resource characterization and fore- casting, like chemical weather monitoring and forecasting, requires infor- mation about vertical structure of mean and turbulent wind characteristics and temperature throughout the atmospheric boundary layer, including boundary-layer depth. A 1 percent error in wind speed characterization has an estimated $12 million impact on projected output of a 100 MW wind-power plant over its lifetime. Variability and uncertainty of near-term (diurnal cycle) and long-term future power deliverable from wind farms underscores the need for vertical profiles of relevant variables at a frequency exceeding twice-daily raob schedules. In the energy sector, weather information translates directly into profits and losses on short time scales (minutes to days). Sensitivity of energy demand to climate fluctuations is illustrated by the fact that a fraction of a degree in

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP Vignette: Duke Power Background Duke Energy­Carolinas generates, transmits, and distributes electrical energy to customers throughout western North and South Carolina. Damaging ice storms in 1996, 2002, and 2005 cost Duke millions in restoration dollars. The most severe ice storm in December 2002 affected a large section of the Carolinas Service Area, caused 1,375,000 customers to lose power, and cost the company $77 million in repairs. That storm resulted in the mobilization of over 11,000 workers from Duke Energy and external companies to repair 3,200 damaged poles, 2,300 transformers, and 549 miles of wire. Because of the huge economic impact of damaging ice storms, power companies take a proactive approach to forecasting, planning, and scheduling resources ahead of such events. Millions of dollars in resource decisions are made before and during an ice storm. Since ice storms are exceptionally damaging and difficult to forecast, real­ time mesoscale observations are critical. Two factors affect ice accumulation on trees and power lines: total rainfall and surface temperature (i.e., how far below freezing). Spatial patterns of rainfall and subfreezing temperatures are important in estimating the ice thickness and areal coverage and affect resource decisions for utility management. One recent event underscores the importance of real­time mesoscale observations in making quick resource decisions. 3-1.eps December 2002 ice storm in the Western Carolinas. SOURCE: Nick Keener, Duke Energy. bitmap image

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS The Event Numerical forecast guidance from the National Centers for Environmental Prediction (NCEP) indicated the potential for a significant ice storm on February 1, 2007, across portions of the Duke Energy­Carolinas Service Area. Forecast models indicated the potential of 3/8 to 1/2 inches of ice accumulation beginning in the early morning hours and continuing into the afternoon of February 1. As early as January 29 Duke Energy mobilized its internal workforce and began contacting neighboring utilities and contractors for additional resources in anticipation of sig­ nificant outages across the service area. During the pre­dawn hours of February 1, freezing rain begin to fall across the area. By mid­morning a light glaze of ice approaching 1/8 inch had accumulated. Radar and forecast guidance indicated that freezing rain would continue well into the afternoon and, if temperatures remained below freezing, would result in ice accumulations approaching ½ inch on trees and power lines before ending. By 10 am on the morning of February 1, Automated Surface Observing System (ASOS) observations showed temperatures rising to near the freezing point. By 11 am the surface temperatures in most affected areas were 32°F. Although the latest numerical guidance at that time and the official forecast from the National Weather Service (NWS) still indicated a continuation of freezing rain, it was apparent that milder air was mixing to the surface from aloft and was eroding the shallow wedge of sub­freezing temperatures. Given the real­time ASOS observations supplemented by other surface reports, Duke Energy’s meteorology staff recommended that preparations for significant outages be discontinued, resulting in a significant savings to the company. Real­time mesoscale observations of temperature, dewpoint, wind, pressure, and precipitation are critical to decision making when a temperature change of a few degrees can mark the difference between a cold rain and an ice storm. The former is a nuisance; the latter causes major disruptions. An increase in the spatial coverage of the mesoscale observational network and the availability of 15­minute reports would enable electric utilities, municipalities, and public transportation entities to make better resource decisions, thereby saving money and possibly mitigating impacts from adverse weather conditions.

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP mean temperature for planning over a heating season can have huge impacts on profitability. Also, a particular electrical power company will do a daily market analysis for the entire nation (i.e., outside its own service area) and so will need reliable observations and forecasts outside its service area. Current Assets and Operational Requirements for Mesoscale Observations The power-generation industry uses a combination of National Oceanic and Atmospheric Administration (NOAA) products and services together with company-managed observations and in-house forecasting. Company- managed observations generally are intended to increase spatial resolution or timely access rather than to introduce new sensing technologies or to observe different meteorological variables. Below are some examples of measurements and forecast time horizons appropriate for national energy security. Stationary Power Generation • Current observations and short-term forecasts for meeting daily load demand, hourly pricing, and risk analysis for load obligation; • Current observations and short-term forecasts of extreme condi- tions leading to possible interruptions and storm response; • Current observations and short-term forecasts of high concentra- tions of pollutants near the ground as a reason for curtailing operations at coal-fired power plants; • Climatological databases and seasonal forecasts for monthly and seasonal planning for fuel inventories and revenue projections; • Climatological databases, and seasonal and interannual to inter- decadal projections of climate variability and change. Long-term planning for urban and rural development, and industrial and commercial needs (climate scales). Biomass Primary Energy Production • Seasonal climate forecast or outlook for biomass crop choice; • Current observations and short-term forecasts for planting condi- tions (soil moisture, soil temperature); • Seasonal forecasts for projecting crop growth and harvested biomass. Renewable Energy Generation • One-to-three-day estimates of atmospheric turbidity and cloudiness for projections of solar power generation;

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS • One-to-three-day estimates of wind speed for estimates of power generation by wind turbines. In the same time frame, icing on the props must be anticipated; • Very short-term forecasts of sudden wind shifts or changes in speed. Turbines must be shut down to avoid damage if wind speeds become too high. Air density forecasts are useful in that density affects the force exerted on props when the wind blows. Needs for the Future Enhanced observations to meet the needs of the power-generation industry would lead to more accurate initial conditions in model forecasts. Similarly, high-resolution observations needed by the Department of Home- land Security (DHS) for running plume models and coordinating emergency response in urban environments would find profitable use on a 24/7 basis in the energy industry. Increased attention to energy security under conditions when foreign sources of primary energy are supplanted by national sources, especially more weather-vulnerable sources such as biomass, will call for more inten- sive weather monitoring and forecasting (short-term and seasonal). At the same time more energy-related measurement platforms may be available for providing new measurements (e.g., wind farm towers, instrumented agricultural machinery, or “smart” power transmission line poles in remote areas). More fine-scale information is needed in urban centers. Distinguishing the meteorological factors (temperature, humidity, wind, solar radiation) that influence energy send-out by urban divisions (e.g., residential, indus- trial, commercial, recreational) would facilitate estimation of short-term demand. Real-time surface data should be available from NOAA Port or its equivalent every 15 minutes. Sites that are able to operate and transmit data through outlier events such as hurricanes and severe thunderstorms are critical for damage assess- ments prior to restoration of power, normal communications, and safe travel in the affected areas. Multiple options for communicating the data to a central facility often mark the difference between the loss of vital informa- tion or continued reception. More detail about the spatial distribution of total rainfall over indi- vidual watersheds is needed to manage reservoir storage. In winter and spring, basin-by-basin data on snowpack and the rate of melting enable decisions about reservoir storage and clarify the potential for flooding on large rivers, which affects inland marine traffic. Both remote sensing systems and in-situ observations are needed to characterize the planetary boundary layer over the diurnal cycle. Knowledge

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP of the depth of the atmospheric boundary layer is needed to operate con- ventional power generation facilities and wind parks. Detailed observations aloft of temperature, humidity, wind, and suspended hydrometeors help to inform resource estimates (wind parks), emergency operations (power out- ages), and curtailment of operations (because of air pollution). Any observations that would improve extended range (seasonal) fore- casts (sea-surface temperatures, temperatures within the thermocline, soil moisture, snowmelt) would immensely benefit the power-generation indus- try, particularly in the production of biomass fuel. A national network of soil-moisture measurements (need not be spa- tially uniform) would benefit several participants in the energy sector. The role of soil moisture as a climate memory for short-term and seasonal fore- casts is well known. Such a network would have a multiplicative benefit, since better observations of soil moisture would improve short-term pre- cipitation (and hence short-term soil moisture) forecasts, which, in turn, would improve seasonal precipitation and soil-moisture climate forecasts. In addition, a network of soil-moisture measurements would benefit stream- flow forecasts, snow-depth forecasts, biomass production estimates, and reservoir-level estimates. When seasonal forecasts achieve a level of accu- racy and quantifiable uncertainty such that they are used routinely in eco- nomic projection models, they have the potential to substantially mitigate negative economic impacts of extreme weather events. One variable not measured by NOAA or other standard networks is the temperature of water discharged from power plants and of adjacent lakes and downstream rivers. Human-caused alterations in natural water temperatures have strong effects on downstream ecology. Influences of weather on the power-generation industry are event driven. Building upon the analysis provided by Schlatter et al. (2005) and follow- ing the methodology we used develop Appendix A, we can summarize the spatial and temporal scales of influences of weather on the power industry (Table 3.1). We can also estimate the measurement resolution (instrument accuracy, spatial resolution, and temporal resolution) needed to meet the needs of the power industry. PUBLIC HEALTH AND SAFETY Importance to the National Economy Safety and health concerns extend beyond the traditional weather issues related to transportation, severe storms, and energy to the important issue of air quality. The chemical composition of the atmosphere has been (and is being) significantly perturbed by emissions of trace gases and aerosols associated with a variety of anthropogenic activities. This changing of

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS TABLE 3.1 Spatial and temporal scales of several meteorological phenomena of consequence for the power-generation industry, and the measurement resolution (instrument accuracy, spatial resolution, and temporal resolution) required to adequately observe those phenomena Time Measurement Resolution Event Space Heat wave (temp) 500-1500 km 2 days-1 week 0.5°C, 10 km, 1 hr 1 m s–1, 1 km, 1 min Winda 1-2000 km 1 min-4 days 0.5 m s–1, 100 m, 10 min; Wind (for wind 100 m-1000 km; 10 min-1 week (1 m s–1, 30 m, 10min)b to 1 kmb power) Snow and ice 50-1000 km minutes-2 days 1 mm snow water equiv. storms 1 cm snow, 1 km, 30 min Lightning region minutes to hours location to 0.5 km Precipitationc basin to regional Hours-days, 1 mm, 1 km, 1 hr. seasonal to interannual Cloudinessc local to regional daytime hourly 0.1 sky, 10 km, 20 min to monthly Waste heat impact 10 km, lakes and 1 hour-4 days 0.5°C, 100 m, 1 h rivers Normal weather urban (2 km) rural 20 min-climate (30 km) aCould be associated with a Nor’easter (4 days), icing conditions, hurricanes or tornadoes (1 min), straight-line winds, or fire weather. bMeasurements in the vertical direction. cCould be from short-term (management) or long-term (planning) for hydropower production. SOURCE: Derived from Schlatter et al. (2005). the chemical composition of the atmosphere has important implications for urban, regional, and global air quality, and for climate change. In the United States, 104 counties are currently in non-attainment with respect to the 8-hour National Ambient Air Quality Standards (NAAQS) standard for ground-level ozone (Figure 3.1). The situation is expected to worsen as we move towards even more stringent standards. Safety and health concerns also extend beyond traditional air quality issues to encompass the effects of heat waves, severe cold, and high pollen levels, and to emergency response to release of hazardous substances, bioterrorism, and fires/smoke.

OCR for page 42
0 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP FIGURE 3.1 U.S. counties in non-attainment of the NAAQS 8-hour ozone stan- 3-2.eps dards. SOURCE: Scheffe (2007). bitmap image Dealing with public health and safety issues requires an ability to char- acterize and predict chemical weather. By chemical weather we mean “local, regional, and global distributions of important trace gases and aerosols and their variability on time scales from minutes to days, particularly in light of their various impacts, such as on human health” (Lawrence et al., 2005). As in the field of meteorology, prediction involves both observations and models, and their close integration. The use of chemical weather forecasts in Public Health and Safety (PHS) management has become a new application area and provides important information to the public, decision-makers, and researchers. Many cities in the United States are providing real-time air quality/chemical weather forecasts and various organizations are broaden- ing their services to include prediction of other environmental phenomena (e.g., plumes from biomass burning, volcanic eruptions, dust storms, and urban air pollution) that could potentially affect the health and welfare of their inhabitants. For example, the National Weather Service (NWS) has

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS recently started to provide mesoscale numerical model forecast guidance for short-term air quality predictions, beginning with next-day ozone forecasts (available at www.weather.gov/aq), and plans to expand this air quality capability by extending the forecast period and adding fine particulate matter (PM2.5) to the forecasts. Borrowing lessons learned from the evolution of numerical weather prediction (NWP), chemical weather prediction through the assimilation of chemical data holds significant promise. Careful design and use of observa- tions will produce an expanded capability in chemical weather prediction, which in turn will offer benefits in the following areas: • Public health: Accurate time- and location-specific health alerts will help the public reduce acute exposure when high pollution levels are expected. Routine daily forecasts will enable the public to make healthier choices (e.g., exercising outside only on low-pollution days). • Planning: Chemical weather forecasts will allow organizations to plan business activities more effectively. For example, the U.S. Forest Ser- vice and other land management agencies will need these forecasts to ensure their planned ten-fold increase in prescribed burning will not result in viola- tions of the NAAQS. Forecasts can be used by government and industry to reduce emissions on predicted high-pollution days, thus avoiding the high cost of continuous emission controls. • Emergency response and risk management: Effective emergency- response forecasting will help organizations better understand and manage the consequences of accidental or intentional releases of hazardous material into the atmosphere. With that information, they can reduce exposure, both by effective responses (e.g., sheltering-in-place, evacuating) and by planning remedial actions. • Forensics: Identifying the type and quantity of hazardous materials released into the atmosphere will require not only measurements but also accurate dispersion modeling of plume concentrations and ground deposition. • Wildfires and smoke: Improved prediction of chemical weather will assist air quality agencies in planning controlled burns, as well as aiding firefighters in setting up command posts, managing or fighting fires, and protecting themselves from exposure to smoke. Additionally, the public will benefit from evacuation guidance and protective measures. • Assessments: Chemical forecasting simulations and their reanalysis will provide valuable continuous records of air quality and deposition estimates that will inform numerous retrospective assessments such as epi- demiological studies, the progress of air program rules, and delineation of meteorological and emissions influences on air quality.

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP Current Assets and Operational Requirements for Mesoscale Observations Meteorological Parameters Air quality and related issues depend strongly on the meteorological conditions that effect dispersion and emissions (e.g., wind-blown soil fluxes depend on surface winds, and evaporative emissions depend on tempera- ture). Thus a better characterization of meteorological conditions directly benefits air quality prediction and management. However, there are some important differences in the spatial and temporal requirements for air quality mesoscale observations. Many PHS problems are associated with benign weather (stagnant conditions) and particular phenomena such as urban heat islands, nocturnal jets, local circulations (e.g., sea-land breezes), which have not been the primary focus of observing systems nor NWP efforts (which have been slanted more toward severe storm conditions). Boundary-layer information, including mixing layer height and clouds, is of particular importance. Key meteorological parameters for PHS applications include temperature, wind speed and direction, boundary-layer character- ization, relative humidity, and solar radiation, often at scales of less than a kilometer horizontal spacing (i.e., city-block scale). Pollutant Parameters There are observational needs beyond those of meteorology. Measure- ments of key trace gases and aerosols are required in PHS applications. Many observational networks focus on air quality and related issues and are too numerous to list here. A sampling of air quality networks is pre- sented in the second table of Appendix B Table B.2. Many are operated by the Environmental Protection Agency (EPA). The National Park Service (NPS), NOAA, the Department of Energy (DOE), and the U.S. Forest Service (USFS), state agencies, and tribal governments are involved in the operation of air monitoring networks, and some networks are privately operated by interested industry or research groups (Scheffe, 2007). In addition to the networks summarized in Appendix Table B.2, there are a number of environmental networks comprised of monitors deployed at power and industrial facilities for compliance and other purposes. Examples include those operated by the Tennessee Valley Authority and the Electric Power Research Institute. These networks are developed to meet fairly specific objectives along programmatic lines. Examples include tracking trends of acidity and acid-neutralizing capacity through the surface water Temporally Integrated Monitoring of Ecosystems /Long-Term Monitoring (TIME/LTM) networks, determining compliance with the (NAAQS) in the State and Local Air Monitoring Networks (SLAMs), and establishing vis-

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP recharge of the groundwater. The time scale of the short-term storage ranges from hours to weeks in rivers and months to seasons in lakes. The long-term storage is replenished at very slow rates but results in residence time on the order of years. On the consumption side, the distribution of the population density, industrial and agricultural activities, and the tradeoffs between the needs for flood control, energy production, and recreation con- trol water availability or shortage. While certain activities (e.g., municipal, industrial) result in little water loss, others, such as agriculture, require large quantities with significant losses to the atmosphere. The “big picture” of water quality is even more complicated. Indus- trial and municipal activities result in water pollution, and the need for water treatment. This treatment is never complete, and natural processes are counted on to assist in the treatment. The plethora of chemicals used in industrial processes prevents comprehensive and complete addressing of their environmental impact. As a result, surface and ground waters contain heavy metals, toxins, pharmaceuticals, and deadly bacteria. As water is a major transport agent, the effects extend thousands of miles in spatial scales and persist in the environment for many years if not biodegradable. Rainfall and agricultural practices result in soil erosion, and the displaced particles carry with them both fertilizers and pollutants. Over time these travel to streams and rivers and affect environments thousands of kilometers away. A prime example here is the problem of hypoxia in the Gulf of Mexico, which traced back to nutrients used in the Midwest for crop production. This brief overview of national water resources should be sufficient to illustrate the extremely complicated nature of the problem. The processes of water supply, storage, and consumption span a tremendous range of scales both in time (from seconds to decades), and in space from (sub-millimeters to thousands of kilometers). The data and other information about these processes are scattered across many organizations and are difficult to access in a comprehensive fashion despite efforts such as the National Integrated Drought Information System (Western Governors’ Association, 2004). Sig- nificant gaps exist in our knowledge about the functioning of the natural water systems. Current Assets and Operational Requirements for Mesoscale Observations Addressing the problems of “too much water or not enough” requires careful monitoring, skillful prediction, and rational control. Federal respon- sibilities are organized along those lines with some degree of overlap. For example, the United States Geological Survey and the Bureau of Reclama- tion monitor surface and groundwater status in terms of quantity and quality using information from some 1.5 million sites, EPA enforces compli- ance with environmental standards and regulations, and NOAA focuses on

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS the coastal waters and the atmosphere. The National Weather Service has the mandate to routinely forecast stream flow at some 20,000 points across the country. Toward this end, it has established the Hydrometeorological Automated Data System (HADS)17 to collect raw hydrological and meteo- rological data from sites operated by a variety of agencies, using the Data Collection Platforms aboard GOES satellites. The U.S. Army Corps of Engi- neers develops, monitors, and operates engineering structures such as reser- voirs, and dams and locks on major rivers. In many regions of the country these responsibilities are shared with state and other local agencies. For example, Florida’s water resources are managed by four water management districts that are responsible for providing water for municipal and rural consumption, agricultural use, and ensuring protection of life and property. The Tennessee Valley Authority operates numerous water storage reservoirs for water supply, flood control, and electric energy production in the South- east. Comprehensive discussion of all the federal, state, and local agencies and their activities related to the nation’s water resources is prohibitively complex and beyond the scope of this report. These complexities are also emphasized by an American Meteorological Society policy statement on water resources (AMS, 2008). Since there is no single agency responsible for management of water resources, and there is no national water policy (Galloway, 2006), our discussion will focus on the major mechanism of observational capabilities that we consider critical for addressing multiple national needs. Serving multiple national needs requires decision-making regarding water apportioning and restricting. These decisions often address the con- flicting objectives of different users and are made using incomplete informa- tion regarding the current and future state of water resources of interest. To support this decision-making, responsible parties use predictive models that interpolate and extrapolate the available data into the variables of inter- est that are often not observed directly. Examples of such models include rainfall-runoff transformations, flash-flood forecasting, flood routing along main rivers, groundwater recharge and flow, land-atmosphere interaction with estimates of evapotranspiration, sediment transport and sediment yield, snowmelt, and water storage, just to name a few. These models are highly uncertain because they describe complicated nonlinear processes that are difficult to observe and that are the result of multiscale interactions of other processes. The predictive skill of these models varies, but uncertainty is an inherent part of any of them. This uncertainty, which is hard to quantify, can be attributed to (1) the lack of complete understanding of the processes involved, (2) suboptimal parameters of the mathematical representations that constitute the models, (3) errors 17See http://www.nws.noaa.gov/oh/hads/

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP in the initial conditions, and (4) errors in the main input (e.g., rainfall in the flash-flood forecasting models). Much of the above uncertainty can be ascribed to the limitations of our observing systems, both for opera- tions and for research, that is, for learning about the processes of interest. These observing systems provide the empirical information that is used to express formally our understanding of the natural processes, calibrate (i.e., adjust the parameters of) our models, and provide the initial condition and the driving inputs. Therefore, the uncertainty due to the limited scope (spatial and temporal sampling resolution and accuracy) of our observ- ing systems propagates all the way to decision-making on the use of our national water resources. For example, Welles et al. (2007) report a lack of significant progress in the skill of stream-flow forecasting models over the past 20 years. Improving the forecast quality by replacing the currently used models with a new-generation of spatially distributed representation of hydrologic processes requires adequate observational input. Studies of the “forecast worth” in the context of reservoir operation clearly show significant economic benefits if reservoir inflows are known more accurately (e.g., Georgakakos et al., 2000). The inflows taken in combination with the current storage determine water availability. This, combined with predictions of water demand, or the demand for energy that can be produced by releasing water through a turbine system, and subject to environmental constraints (e.g., to satisfy the minimum required discharge to sustain ecology downstream), results in a decision on how to operate the reservoir. Needs for the Future The question arises, “What hydrologic model requirements are needed to realize improved prediction relevant to the problem we are discussing?” The answer is not straightforward, and it depends on the spatial and tem- poral scales involved. Consider first a large river. To make a prediction of the discharge at an arbitrary point downstream, we need to know the discharge upstream and an estimate of the inflow into the main channel between the two points. For this we need channel routing models based on the principles of fluid flow in an open channel. The hydraulic characteristics such as slope, width, bottom and bank roughness, and water height deter- mine the answer. When the river basin is large enough, a convective storm, even one with high rainfall intensity, hardly matters, because what happens at the point of interest downstream is mainly affected by the water flow already in the channel. At the opposite end of the spectrum is flash flooding in a small basin. Here, what happens in the channel is largely irrelevant for the forecasting of discharge at the basin outlet, because it will change quickly (within 10-30 minutes, depending on the location and basin size) if

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS enough rainfall occurs. The amount of rainfall and the physical-topographic characteristics of the basin matter most. Among those characteristics it is the level of water storage in the upper zone of the soil that determines the partitioning of rainfall into runoff. This storage can be estimated if mea- surements of the soil-moisture profile are known. Soil moisture plays a major role in several hydrologic processes that affect water resources at many spatial and temporal scales. It controls parti- tioning of rainfall water into surface runoff and the infiltrated water. Surface runoff constitutes the fast-response component by a basin, with water flow- ing through the channel network to the basin outlet. The infiltrated water is either consumed by plants or percolates to deeper storage, recharging groundwater aquifers. Soil moisture also controls the partitioning of energy incident on the land surface. Water availability for consumption by plants leads to evapotranspiration, a major component of the surface energy bud- get. Prolonged water deficit in the root zone affects the life cycle of plants and eventually leads to changes in the surface albedo. Water transported to the atmosphere by evapotranspiration affects thermodynamic processes and, under the right circumstances, precipitates, usually at locations far removed from its origin. Through these processes, moisture in the top layers of soil acts on shorter time scales, to influence day-to-day weather, whereas the amount of soil moisture at deeper levels impacts slower processes at regional scales and acts as a source for water that deep-rooting plants bring to the atmosphere during extended periods without precipitation. Clearly, knowledge of actual soil moisture from a network of stations can help forecast stream flow, evapotranspiration, groundwater discharge, and precipitation. But measurements of the soil-moisture profile some 2 m deep are needed. Moisture in the near-surface layer controls the infiltration capacity and changes rather quickly. The lower layer includes most of the soil and provides moisture for certain class of plants (e.g., grasses). Water is transported between these two layers by gravity, root suction, and capillary suction. The water content in that layer fluctuates more slowly than in the top layer. A still deeper layer extends beneath the soil and provides water for larger plants (e.g., trees). It fluctuates slowly, and its depletion is a sign of a major drought. Spatial variability of soil moisture is high and not very well understood. It is controlled by the variability of rainfall, elevation, slope exposure, land use, land cover, and the hydraulic characteristics of the soil. All these characteristics vary, the last one perhaps the most significantly, because it depends on the pore size distribution and the structure and composition of the soil particles. In the absence of a comprehensive national network of soil moisture observations, our understanding of this variable is based on experimental in-situ data (Illston et al., 2008), focused remote sensing campaigns, and modeling studies.

OCR for page 42
0 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP Currently available remote sensing technologies cannot provide moisture observations for the entire soil column. The signal measured by radiometers deployed on aircraft or spacecraft originates from roughly the top 5 cm of the soil and is often obstructed by the water content in and on the plants. The spatial resolution of microwave observations of soil moisture is a func- tion of the frequency, antenna size, and the antenna range. While aircraft- mounted radiometers can provide data with the resolution of about 1 km or better, national coverage requires satellite-based sensors. This translates into resolution on the order of 5 km (Entekhabi et al., 2004). Methods of in-situ measurement of soil moisture take advantage of various physical phenomena (Raats, 2001). Perhaps the most practical method is Time Domain Reflectometry, in which the propagation velocity of an electromagnetic pulse, which depends on the water content, is mea- sured. While these probes require careful calibration, they are inexpensive, safe, and easy to install. Any soil-moisture measurement network should also include soil-temperature sensors. These easy to make measurements would help predict surface runoff when rain falls on frozen soil, helping to mitigate effects of frequent spring flooding in parts of the country. Recommendation: A national, real-time network of soil moisture and soil temperature observations should be deployed nationwide at approximately 3000 sites. This number corresponds to a characteristic spacing of approximately 50 km for a network that is spatially distributed across the continental United States. Although this spacing is insufficient to capture the full spectrum of short-term spatial variability of surface soil wetness, it is small enough to represent seasonal variations and regional gradients, thereby supporting numerous important applications such as land data assimilation systems in support of numerical weather prediction, water resources management, flood control and forecasting, and forestry, rangeland, cropland, and ecosystems management. This characteristic spacing would also provide data at a resolu- tion that complements historical and relevant datasets. Site selection should be biased toward existing networks, provided that the instrument exposure is acceptable and real-time communication is possible. Although we argue for the deployment of a national network of soil moisture measurements to improve the prediction of water movement on the surface and below, precipitation remains the most significant variable that determines runoff. Currently two major sensors are used to monitor precipitation: rain gauge networks and weather radars. In principle, the combined use of these two systems should provide detailed and accurate depiction of rainfall across the country. Unfortunately, this is not the case, especially at the short time scale relevant to flash-flood prediction. Partly

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS due to very high spatial and temporal variability of precipitation, and rain- fall in particular, the accuracy of rainfall maps is not very high. Ciach et al. (2007) report random errors approaching 50 percent for hourly rainfall maps produced by the national network of WSR-88D weather radars, also known as NEXRAD. As the temporal scale of rainfall accumulation increases, the random errors decrease, and, as a result, seasonal maps of precipitation depict correct pictures of the process. The major problem is sampling. Rain gauges are distributed too sparsely to capture the variability of rainfall patterns, in particular those of convec- tive origin. Radar beams “look” slightly upwards and tend to overshoot clouds at a certain distance. Radars located on mountaintops in the West consistently miss precipitation originating in clouds at lower elevations. A solution is to place small, inexpensive radars that survey relatively small domains (~1000 km2), such as urban areas or sections of mountainous terrain. We discuss such systems further in Chapter 4. Another variable that is not observed as densely as it should be is the stream flow. The USGS monitors the nation’s rivers in real time at some 1700 sites. Other agencies complement this at the sites they operate. This is not sufficient coverage, considering the complexity of water movement though the landscape and the primary role that water transport plays in other biogeochemical processes. Continuous observations of stream and river discharge provide a dual benefit for the general problem of forecasting and control of water resources. On the one hand they are directly relevant to flood forecasting and reservoir inflow prediction, and on the other hand they provide a constraint on the models used to forecast other elements of the water cycle that are critical for numerous applications. These include groundwater discharge, pollution transport in surface and ground water, and evapotranspiration. Accurate monitoring of the major fluxes and storages of the water cycle is a prerequisite to improved prediction of many environmental problems affecting the nation. Transport of sediment originating in soil erosion is a result of agricultural practices and vegetation, erosive power of rain- drops and wind, concentrated surface runoff, and transport along the river channel network. Sediment carries both nutrients and pollutants that are attached through cohesion and undergoes transformations while traveling. It affects many other quality aspects of the surface waters and their biologi- cal environment by changing the turbidity and acidity. Current techniques for stream-flow estimation are expensive and labor intensive. Building the structure that houses the sensor that measures stage (depth) is the major expense. The relationship between stage and discharge is developed empirically by periodic and more direct measuring of the discharge, as a product of water velocity and channel cross-section. The empirical data collection has to be repeated over time so that a full range

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP of variability is represented. For high flows this presents practical difficulties and risk to the crew. Recent developments include advanced contactless technologies, and several are being tested and researched. The techniques range from optical sensors to active remote sensing using low-power radars. Other approaches involve computational fluid mechanics models to develop the rating curve and an inexpensive stage sensor for converting to discharge. Some of these techniques are inexpensive and could complement the core networks oper- ated by the USGS and other agencies. The observational limits of various aspects of our nation’s water resources have been recognized by the research community. Hydrologists and environ- mental engineers argue for the development of a network of well-instrumented natural observatories to further our understanding of water movement in the environment. Comprehensive observations of water quantity and quality are required to improve our predictive capabilities to benefit society. Details of the arguments are provided in CUAHSI (2007) and WATERS (2008). FOOD PRODUCTION Importance to the National Economy Food is grown in all regions of the United States, with each region tak- ing advantage of local climate and soils to exploit its competitive advantage for specific food-related products. The relatively inexpensive transporta- tion of the 20th century has reduced the incentive to raise a wide range of food crops, including some that are only marginally adapted to local soils and climate, in every region. With inevitable rises in transportation costs and increased interest in locally grown food, future weather and climate information needs for food production may be more extensive than in the past. Fruits and vegetables grown at the margins of their optimal ranges are more vulnerable to influences of drought, flood, water-logged soils, heat stress, cold stress, cloudiness, too high or too low humidity, diseases, insects, herbivores, growing season length, or other factors relating directly or indirectly to climate. The commodity crops of corn, soybean, wheat, oats, barley, rye, etc. are grown on vast areas as monocultures. Food consumed by humans in the categories of fresh fruits, nuts, and vegetables, by contrast, are con- sidered relatively high-value crops and are grown in smaller plots with higher income per unit area, more intense use of labor, more frequent use of irrigation, and higher costs of production per unit area. For various rea- sons, monitoring meteorological conditions on smaller spatial and temporal scales may be more important in regions growing specialty crops than those growing commodity crops. This will become increasingly so as specialty

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS crops are raised more widely in climatically marginal regions or regions commonly devoted to commodity crops. Animals are raised in every region of the United States for meat, milk, and egg production. But like commodity grains, commodity meat produc- tion (beef, pork, poultry, fish) tends to concentrate in certain areas where there is access to feed grains, abundant water, optimal temperature or pre- cipitation regimes, and access to transportation or markets. Animals used for meat, milk, and egg production may be raised in confined spaces (either indoor or outdoor) or in “free-range” environments. Confinement operations present additional environmental issues, such as high volumes of odor dust and waste, that call for additional environ- mental monitoring. Free-range (i.e., grazing) operations typically cover large areas where growth and productivity of grazing materials are factors to be monitored. Extreme events such as heat waves, freezing rain, extreme cold, severe storms, excessive rain or snow, or high humidity can have serious impacts on animal productivity or even mortality. Weight gain in meat animals, egg production, milk production, and the success of animal breeding are all negatively impacted by extreme high temperatures. Cold rain followed by sub-freezing temperatures leads to sickness in beef animals raised with- out shelter. The monitoring of current conditions and access to reliable short-term forecasts would allow pre-emptive actions to minimize adverse weather effects on animals raised under confined conditions. U.S. food production feeds a population of over 300 million people. In 2007, the United States exported about $82 billion in commodity crops, livestock, and horticultural products. In 2008, the total is expected to be $114 billion.18 Current Assets and Operational Requirements for Mesoscale Observations Environmental conditions monitored for agricultural crops usually include standard surface meteorological variables but also include photo- synthetically active radiation (PAR), evapotranspiration, soil temperature, and soil moisture. For some crops, leaf wetness (as a measured variable) is a critical factor for management decisions relating to pests and pathogens. Of these variables, the one least likely to be observed, and yet of critical impor- tance for many regions, is soil moisture. The heterogeneity of soils and land- scapes make representative observations of soil moisture a major challenge. The increased education and sophistication of agricultural producers, coupled with the increased availability of weather and climate information 18 U.S. Department of Agriculture, available at http://www.fas.usda.gov/cmp/outlook/00/ Aug-0/AES-0--00.pdf.

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP over the Internet, has intensified use of such information by producers and by agribusiness service providers for near-term management decisions, long- term plans for marketing, investments in conservation practices, and water management (irrigation, tile drainage, grass waterways). Modern farm machinery comes equipped with devices for measuring and recording plant- ing rates, chemical application, and grain harvest yield, all as a function of the (high-resolution) position in the field. This high spatial detail, along with the high spatial detail of weather and soil conditions begins to reveal previously unavailable opportunities for maximizing yield and reducing adverse environmental impacts. Agribusiness service providers also must be knowledgeable of current and future weather conditions for maintaining material inventories, managing storage facilities, and generally anticipat- ing weather-driven demands by producers for their goods and services. The crop insurance industry has a major interest in reliable and accurate weather and climate information, especially under potentially changing climate, and in extreme events such as high wind, tornadoes, drought, hail, and freeze occurrences. Needs for the Future Weather Data for Driving Decision-Support Tools The increased use of decision-support tools in agriculture based on current or projected future conditions calls for a wider range of measure- ments, higher density of sensors, and higher frequency of observations. High-density surface wind observations go into the formula for estimating evapotranspiration and determine when conditions are favorable for apply- ing pesticides or starting controlled burns in the fields. Models of growth for commodity crops use past, current, and future predicted weather and allow producers to plan management and marketing activities. Decision- support tools can be designed to alert producers of impending disease or insect outbreaks when future conditions favor such events. Examples of methods used to increase profitability or environmentally sustainable agri- culture include decision-support tools and models for predicting soil ero- sion, nitrate leaching, soil moisture, soil temperature, irrigation scheduling, forage quality, sub-surface drainage tile flow, stream-flow, water quality, insect migration or infestation, fungal growth, milk production, and weight gain in meat animals. The storage of grain and transport of both grain and animals to market are vulnerable to weather-induced hazards or reduction in product quality.

OCR for page 42
 NATIONAL NEEDS FOR MESOSCALE OBSERVATIONS Bio-Economy and Increased Needs for Weather Information National mandates for the increased use of biomaterials to replace fossil fuels for mobile transportation have intensified the need for enhanced biomass production from agricultural lands. Future needs to raise increased amounts of both food crops and fuel crops from the same land area will heighten the role of weather and weather forecasting—especially seasonal forecasting—in decision making in the bio-economy. A wider variety of crops in regions now commonly dominated by monocultures of commod- ity crops will experience changed surface-atmosphere interactions, such as changed evapotranspiration, which in turn could alter the precipitation recycling ratio. Interannual variability in cropping choices therefore could contribute an anthropogenic component to the interannual variability in regional climate. Bringing marginal land into production because of the profit incentive of higher commodity prices for feed grains and biofuels may require special monitoring; such lands are marginal because they are highly erosive or are located at the margins of cropping regions due to their soil or climate conditions. Soil storage of carbon is emerging as a method of sequestering carbon from the atmosphere. Microbial processes in soil that regulate the conversion of labile carbon to carbon dioxide are highly temperature and moisture dependent, suggesting a need to monitor these conditions as a means of monitoring carbon storage. All these factors raise the urgency for higher density meteorological and soil measurements for the bio-economy. Water Quality Observations Related to Food Production Emerging issues of surface water quality (with negative contributions by chemical-laden runoff from agricultural lands), long-term sustainability of agricultural practices, and soil sequestration of carbon to meet the goals of reducing concentrations of atmospheric carbon dioxide likely will increase the demand for additional environmental measurements. Surface- water-based measurements of temperature, stream flow, dissolved oxygen, particulate loading, nitrate and phosphate concentrations, and pesticide concentrations are of most interest. By following the analysis provided by Schlatter et al. (2005), we can estimate the spatial and temporal scales of influences of weather on food production (Table 3.3). We can also estimate the measurement resolution (instrument accuracy, spatial resolution, and temporal resolution) needed to meet the needs of the various food production areas (some are speculative and require validation).

OCR for page 42
 OBSERVING WEATHER AND CLIMATE FROM THE GROUND UP TABLE 3.3 Spatial and temporal scales of several meteorological phenomena of consequence for agricultural industries, and the measurement resolution (instrument accuracy, spatial resolution, and temporal resolution) required to adequately observe those phenomena Measurement Event/variable Space Time Resolution Heat wave 500-1500 km 2 days-1 week 1°C , 10 km, 1 h (temperature) Drought (soil 500-1500 km 2 weeks to 2 mm moisture) interannual 1 m s–1, 1 km, 1 min Wind 1 km-2000 km 1 min-4 days Precipitation 10 km-regional hours to days 1 mm, 1 km, 1 h seasonal to interannual Cloudiness local to regional daytime hrly to 1.1 sky, 10 km, 20 min clim Temperature 500-1500 km seasonal 1°F, 10 km, 1 h Flood 0.1 km-100 km 2 days-2 weeks sub-basin Hail 0.1 km-20 km 5 min-5 h 100 m SOURCE: Derived from an analysis provided by Schlatter et al. (2005).