Presentation on theme: "Climate Observations in Pacific Northwest National Parks Kelly T. Redmond Western Regional Climate Center Desert Research Institute 2215 Raggio Parkway."— Presentation transcript:
Climate Observations in Pacific Northwest National Parks Kelly T. Redmond Western Regional Climate Center Desert Research Institute 2215 Raggio Parkway Reno Nevada voice fax
National Weather Service Cooperative Network Approximately 5000 daily max/min temperature stations, 8000 daily precipitation stations, 3000 automated hourly precipitation stations.
Volunteer Observers. Standard raingage 8” wide, 20 inch capacity, This one with funnel (summer configuration). Max/Min Temperature System (MMTS) in background (electronic thermometers). Longest individual observer: 76 years, Cottage Grove Oregon (not the person pictured).
344 U.S. Climate Divisions.
National Park Service Protocol Considerations (stemming from CHIS) Why ? - Why are the measurements being made? Immediate project needs (a few months to a few years duration) “General purpose” - often turn out to be very important & heavily used. Should assume this as the default case. Data from many networks are being used for far more than their original purpose or motivation Sustainability - Acquisition and deployment are the easy and cheap part. Think ahead: Arguments for equipment is usually more effective than arguments for maintenance, quality control, archive, access, summary. Maintenance – the bane of all systematic measurement programs, and (the lack thereof) often responsible for the death of measurement efforts Automated systems still require considerable human attention. Automation is not a panacea, or a replacement for people. Automation introduces new types of errors.
What constitutes a good climate record? - One that records real climate variability rather than fake climate variability. Real Climate Variability: All variations seen in the record are due to things that happened in the atmosphere. And, we usually hope, in the larger scale atmosphere. (Though not always … depends on the purpose of the measurements.) Fake Climate Variability: Sensor changes Replacements, degradations, drift, insects, birds. Site changes Ground cover and land use changes Obstructions / blocking Anything affecting the local energy balance Observer and methodological changes How observations are recorded (conventions followed?) a.m. / p.m. issues. How instruments are reset.
“Climate quality” measurements require a higher standard than “weather quality” measurements. Representativeness - What spatial and temporal scales are the measurements intended to represent? Siting. Choosing a site. What factors, at what spatial scales, will influence the measurements? The scales vary from millimeters to megameters. What is the site representative of? Practical and scientific considerations both matter, can be at odds, and often lead to compromise. Precipitation – Assume that all precipitation will occur in frozen form, even if just occasional. Can the equipment handle this? Link into existing systems Instrumentation, communication, dissemination Archival, retrieval, analysis Local versus external dependencies. Develop or rely on internal mechanisms and expertise? Develop relations with or rely on external mechanisms and expertise?
Quality control begins at the station. This bears repeating. Quality control begins at the station. The best form of quality control is to not let bad data leave the station at all. Up-front attention to equipment quality, installation quality, robustness to disruption, backup systems. The weather is usually the most important reason that weather measurements are lost. The weather that knocks out weather stations is often the weather we most want to know about. (Extremes, and disturbances.) Design for the worst conditions. Once the data values leave the station, essentially only bad things can happen them. “Cheap” is often not cheap. Two-way communication. More desirable. Enables re-transmission of data, resetting of data loggers, better debugging of problems.
Documentation (“Metadata – data about data”) Need history of whatever factors that can affect the interpretation of the measurements. These include sensors, site physical circumstances (including vegetation), observing practices and reporting conventions, numerical pre-processing, averaging intervals, gage shielding, etc. Most perishable: that which is carried in human heads, best retrieved before or retirement, or more final outcomes. Filing cabinets, old diskettes, etc, also contain much interesting material, but often separated from their originator. GPS positions. Full photo sets, repeated periodically, approximately 1-2 years. Typically 8 point compass, wide angle, skylines and foreground, plus any other helpful angles which show local influences, and changes through time. Changes affecting local climate, within meters. Irrigation starting or stopping, vegetative regrowth or trimming, pavement or change in surface status, additions or subtractions of obstructions. Digital versions of the above, accessible with the data.
Observing strategies Is the emphasis on learning about temporal behavior, temporal behavior, or both at once? Dependence on spatial correlation structure of temporal behavior of climate elements. (How well does time-varying behavior correlate in space, at various combinations of spatial and temporal scales?) This varies widely, depending on topographic complexity (height, slope, aspect), proximity to water bodies, as well as latitude and season. *** Catch 22: Chicken and egg! We need the observations to tell us what observations we need! *** Solution: Experience, brains, intuition. A few very good and reliable sites, or lots of mediocre sites? Vulnerability assessment. Where could key failures in measurement result in loss of valuable information or knowledge? Example: manual or simple backup of automated precipitation gages, or, saving of water after initial precipitation measurement (we might lose the timing but know the total). Redundancy is not a weakness. A degree of redundancy is good.
Observing strategies (continued) Station density. Optimum density depends on spatial correlation structure mentioned above. General consideration: are all of the principal climate regimes adequately sampled? A mixed strategy seems best: A few reference quality sites that get lots of attention, produce complete records, have needed backup, and are thoroughly documented, interspersed with more numerous “satellite” observations of perhaps lesser quality, or less protected from their various vulnerabilities. Adaptive monitoring. Smarter electronics have made it possible for observing systems to modify their sensing rate, in response to environmental cues. Systematic measurements are always needed, but these can be interspersed with rates that vary according to the behavior of interest. Of course, we don’t always know what behavior will turn out to be of interest, especially if new. Retrospective capability. We often do not learn that a system has experienced a change until well after the fact. We are therefore dependent on systematic measurement to reconstruct the past.
Added considerations raised during meeting. Public interface needed. Web access to reach broad range of public, research, and other user communities. Public outreach / education. There is a great deal of public interest in climate. Capitalize on that as part of the park mission. Administrative environment. Take it as given that personnel are busy, overwhelmed, have multiple duties. “Expectationless” monitoring. The most useful monitoring has no special hypotheses in mind. Ten commandments for climate monitoring Partnering / leveraging. Why re-invent the wheel? Share resources and expertise. Be careful to pick partners with staying power.
Added considerations raised during meeting (continued). Monitoring the monitoring. Is the monitoring performing as needed or intended? Is it functioning at all? How healthy is this activity? Refered to as "network health”. Snow issues. Many of these, in higher, northern parks. Far more challenging, difficult, expensive, labor intensive, to obtain quality records. High elevation monitoring. Badly needed. High elevations may not vary in time like nearby low elevations. Separate monitoring needed. Needs high standards, rugged equipment, good communication, highly reliable instruments. The world changes. Landscape, vegetation, growth can alter locally measured climate significantly. Fires destroy, then grow backi. Clearcuts: slow changes in small scale climate. Mt St Helens: Destruction, then regrowth. Local temp & precip climate will slowly change (and separately).
Ten Principles for Climate Monitoring For the past few years, with respect to climate data issues, frequent mention is made of a set of climate monitoring principles enunciated in 1995 by Tom Karl, director of NCDC. In some quarters (following Gene Rasmusson, for one), they have also been informally referred to as the "Ten Commandments of Climate Monitoring". Both versions are given here. collated by Kelly Redmond, Western Regional Climate Center, August Cliff's Notes version. "Ten Commandments of Climate Monitoring" 1. Assess the impact of new observing systems or changes to existing systems prior to implementation. "Thou shalt properly manage network change." (Assess effect of proposed changes.) 2. Require a suitable period of overlap for new and old observing systems. "Thou shalt conduct parallel testing." (Of old and replacement systems.) 3. Treat the results of calibration, validation, algorithm changes, and data homogeneity assessments with the same care as the data. "Thou shalt collect metadata." (Full documentation of system and operating prodecures.)
\ 4. Ensure a capability for routine assessments of quality and homogeneity, including high resolution data for extreme events. "Thou shalt assure data quality and continuity." (Assess as part of routine operating procedures.) 5. Integrate assessments, like those of the International Panel on Climate change, into global observing priorities. "Thou shalt anticipate use of the data." (e.g., integrated environmental assessment; anticipate data use as part of operating system plan.) 6. Maintain long-term stations. "Thou shalt worship historical significance." (Maintain long term observing systems which provide homogeneous data sets.) 7. Put a high priority on increasing observations in data-poor regions and regions sensitive to change and variability. "Thou shalt acquire complementary data." (New sites to fill in observational gaps.) 8. Provide network operators, designers, and instrument engineers with long-term requirements at the outset of the design and implementation phases of new systems. "Thou shalt specify climate requirements." (Designers of networks be aware of monitoring requirements for climate usage.)
9. Think through the transition from research observing systems to long-term operations carefully. "Thou shalt have continuity of purpose." (Stable, long-term commitments.) 10. Focus on data management systems that facilitate access, use, and interpretation of weather data and metadata. "Thou shalt provide data and metadata access.“ From Karl et al Full version: 1. The effects on the climate record of changes in instruments, observing practices, observation locations, sampling rates, etc. must be known prior to implementing such changes. This can be ascertained through a period of overlapping measurements between old and new obwerving systems or sometimes by comparison of the old and new observing systems with a reference standard. Site stability for in-situ measurements, both in terms of physical location and changes in the nearby environment, should also be a key criterion in site selection. Thus, many synoptic network stations, primarily used in weather forecasting but which provide valuable climate data, and all dedicated climatolotical stations intended to be operational for extended periods, must be subject to such a policy.
2. The processing algorithms and changes in these algorithms must be well documented. Documentation of these changes should be carried along with the data throughout the data archiving process. 3. Knowledge of instrument, station and/or platform history is essential for data interpretation and use. Changes in instrument sampling time, local environmental conditions for in-situ measurements, and any other factors pertinent to the interpretation of the observations and measurements should be recorded as a mandatory part of the observing routine and be archived with the original data. 4. In-situ and other observations with a long uninterrupted record should be maintained. Every effort should be applied to protect the data sets that have provided long-term homogeneous observations. "Long-term" for space-based measurements is measured in decades, but for more conventional measurements "long-term" may be a century or more. Each element of the observations system should develop a list of prioritized sites or observations based on their contribution to long-term climate monitoring. 5. Calibration, validation and maintenance facilities are a critical requirement for long-term climateic data sets. Climate record homogeneity must be routinely assessed, and corrective action must become part of the archived record. 6. Where feasible, some level of "low-technology" backup to "high-technology" observing systems should be developed to safeguard against unexpected operational failures.
7. Data poor regions, variables and regions sensitive to change, and key measurements with inadequate spatial and temporal resolution should be given the highest priority in the design and implementation of new climate observing systems. 8. Network designers and instrument engineers must be provided long-term climate requirements at the outset of network design. This is particularly important because most observing systems have been desigined for purposes other than long-term climate monitoring. Instruments must have adequate accuracy with biases small enough to document climate variations and chaqnges. 9. Much of the development of new observation capabilities and much of the evidence supporting the value of these observationss stem from research-oriented needs or programs. A lack of stable, long-term commitment to these observations, and lack of a clear transition plan from research to operations, are two frequent limitations in the development of adequate long-term monitoring capabilities. The difficulties of securing a long-term commitment must be overcome if the climate observing system is to be improved in a timely manner with minimum interruptions. 10. Data management systems that facilitate access, use, and interpretation are essential. Freedom of access, low cost, mechanisms which facilitate use (directories, catalogs, browse capabilities, availability of metadata on station histories, algorithm accessibility and documentation, etc.) and quality control should guide data management. International cooperation is critical for successful management of data used to monitor long-term climate change and variability.
Climate Monitoring Guidelines. Sources: Thomas R. Karl, V.E. Derr, D.R. Easterling, C.K. Folland, D.J. Hoffman, S. Levitus, N. Nicholls, D.E. Parker, and G.W. Withee, Critical Issues for Long-Term Climate Monitoring. pp 55-92, in "Long Term Climate Monitoring by the Global Climate Observing System", T.R. Karl, ed, Kluwer, 518 pp. National Research Council, Guidelines and Principles for Climate Monitoring. Appendix F, p 63, in Future of the National Weather Service Cooperative Observer Network. National Academy Press, 65 pp. Eugene Rasmusson, Workshop notes. Climate Services: A vision for the future. NAS/NRC/BASC, Woods Hole MA Dr. Kelly T. Redmond Regional Climatologist/Deputy Director voice Western Regional Climate Center fax Desert Research Institute Raggio Parkwayhttp://www.wrcc.dri.edu Reno, Nevada ftp.wrcc.dri.edu/pub
Magnitudes and perceptions. In a relative sense, climate change is expected to be smaller, perhaps much smaller, than the fluctuations of climate and weather that we already know and love. However, it is also expected to be fairly systematic. Thus, many causes and consequences would likely express themselves as subtle systematic biases in time series of properties of the climate itself and of affected physical and biological systems. These would be interspersed with changes in the likelihood of more visible and memorable occurrences of extremes and disturbances. As cosmologist Philip Morrison pointed out in his book “Nothing is Too Wonderful to be True,” change in the universe occurs approximately equally in quiet and unassuming forms, and in explosive and dramatic forms. But even the latter often build in a slow and innocuous manner.
“Subtle does not mean unimportant.” Prof Mike Wallace UW Atmospheric Sciences Seattle, July 14, 1997