Building capacity and linking infrastructure in the lake and scientific community

Building capacity and linking infrastructure in the lake and scientific community

7-9 March 2005 Scripps Institution of Oceanography - San Diego, California USA

Outcomes from Monday afternoon 7 March

Afternoon session: Breakout group: Lake Observatory Community

Tasks:

List two key science questions that could potentially be addressed from a lake observing system;
List two aspects that may prevent certain scientific goals from being realized.

Coherent themes arising

1) Interactions of physical and biogeochemical processes across time and space scales. How to overlay temporal variability and achieve interpolation to spatial variability?... Are we using the right metric for metabolism and what controls dissolved oxygen across spatial and temporal scales?

- What role do physical factors play in structuring biological (e.g. algal) communities and how this vary across broad temporal and spatial scales

- How does physical limnology control biology

- How do you define lake metabolism in an alkaline lake where HCO3 predominates C uptake?

- Does it mean anything to transfer data before the issue of multiple vs one station in a lake is addressed.

- Spatial variability. How many locations within a lake system do we need to monitor?

- Across scales ranging from second to weeks, what controls dissolved oxygen?

- How do different habitats contribute to whole lake metabolism?

- Can we predict the succession of phytoplankton (species/taxa/biomass) in a lake in response to changes in environment variables?

- Can we decipher physical/chemical//biological signals in an observed biological response.

- What is the role of circulation of biological productivity?

- What is the variability of observed data in the lakes compared to that in terrestrial systems and that amongst continents?

- How does the variation in DOC load (under changing climate) affect lake productivity (P:R ratio)?

2) The role of episodic events in structuring ecological processes;

- What role do physical factors play in structuring biological (e.g. algal) communities and how this vary across broad temporal and spatial scales

- What role do episodic climatic events play in structuring biological communities

- When and where do episodic events occurs that lend to enhanced metabolism (bacterial, phytoplankton, zooplankton) activity? How do these events compare relative to many scale ones to overall ecosystem function? How do we design sampling strategies to capture episodic events?

- Are episodic events an important component of climate change? How do their impacts compare to human influences on the landscape? Can human activities lead to intensified effects of events?

- How do episodic events occurring over relatively short time scales (e.g. internal waves) affect the biological components of lake ecosystems (both short- and long-term)?. What characteristics of lakes constrain the responses to these episodic events?

- What ways of characterizing events provide insight into biological responses across event types?

- The response of lake productivity and C fluxes to extreme weather events (including short-term).

3) Complexity of interdisciplinary models (non-hydrostatic) and the difficulties of calibrating these models, particularly with respect to ecological parameters in physical-biological models.

- How do we gain a greater confidence in biological parameters used in modeling?

- What is the role of non-hydrostatic pressure on lake

4) (Relative importance of global/regional/local drivers of coherence of physical and biogeochemical processes;)

- Coherence on regional/national/global scales of physical/chemical/biological variables. Can we use selected sites to inform us of wider selection of lakes.

- Can we determine which lakes are acting coherently globally, e.g. stratification patterns

- Phytoplankton dynamics: are there common factors which drive the vertical distributions of phytoplankton on a global scale.

- Adaptive sampling: what are the needs for adaptive sampling? How "real time" does sampling need to be?

- Underwater wireless: What are the uses for this?

5) IT issues

- How can we easily accommodate change in sensor configuration/buoy deployment in an automated manner?

- How can we accomplish QA/QC on large amounts of real time data?

- Different questions need different sensors for addressing different lakes

- Stability of data communication under the influence of weather systems?

- Adjustable data schema to automatically accommodate measured changes?

Barriers arising

Logistics/technology

Data management (how to manage, store and access a lot of data?)

Data structure for sensors and tools to generate all subsequent schema changes, QA/QC

Lack of tools, algorithms that do not require manual inspection

Do you have the time scales and adequate data that can deal with the questions of concern?

Lack of spatial/temporal resolution of biological compared with scales of processes

Need for sensors that detect short-term biological responses

Need for algorithms to analyse data streams to detect episodic events?

Methods for sensing biological communities

Reliability - sensor fouling/fault tolerance etc.

Standard spacing of temperature loggers and standard timing of sensors

Fluorescence probes are not in common usage. Would like to distinguish different phytoplankton groups and vertical spacing of probes.

Personnel with truly inter-disciplinary skills (modelers, IT/ecologists)

What are the aspects of modeling that need improvement in order to accurately describe system response?

Non-hydrostatic model may require parallelization?

Once we have a model how can we automate the process of calibration?

Different process descriptions and model set-ups lead to different calibration of parameters.

Good underwater communication merge of data streams, QA/QC of logged data.

Linkages among landscape elements - inexpensive units that capture essential data and which can relay data to convenient notes (e.g. sampling rivers during ice break-up).

Coherent themes arising

2) The role of episodic events in structuring ecological processes;

3) Complexity of interdisciplinary models (non-hydrostatic) and the difficulties of calibrating these models, particularly with respect to ecological parameters in physical-biological models.

4) (Relative importance of global/regional/local drivers of coherence of physical and biogeochemical processes;)

Group 1:

Interactions of physical and biogeochemical processes across time and space scales

To address interactions of physical-biological coupling, what variables would we measure?

Directly measure biological parameters, i.e. types of phytoplankton - different chlorophyll pigments using new fluorometric probes;
Turbidity, suspended sediments;
Temperature/dissolved oxygen;
Meteorological data;
Light – PAR, UV;
Salinity;
Conductivity;
Auxilary data – nutrient concentrations – standardization of analytical methods.

Appropriate Temporal Scale - use the HIGHEST resolution possible – then decide how to present data;

Appropriate Spatial Scales – difficult issue…. How many sites on one lake can we monitor…?

One answer may be to use remote sensing – information on chlorophyll distribution in a lake – modeling – bathymetry…

- otherwise a preliminary detailed study on the horizontal variability… takes time….

Group 2:

The role of episodic events in structuring ecological processes

- At the beginning, a discussion on key processes, driving factors and their diversity; a definition for an episodic event was left open at this stage

- factors causing the event generally meteorological and/or physical:

o wind-induced forcing

o extreme precipitation events

o changes in water level

o timing of ice break-up and spring flood

o changes in land-use (watershed effects: e.g., forest fires, forest clear cut)

o episodic events resulting in a longer-lasting response should also be considered

Research questions:

Question 1: Does an episodic effect lead to a biological or other response?

Question 2: If so, do all lakes respond coherently, and if not (which is rather evident), what is the reason?

- response of an individual lake affected, e.g., by the lake size, other morphometric features, the presence and the type of vegetation in the littoral zone, etc.

Additional questions:

On a local scale, how do the lakes respond to a well identified, individual, episodic event (e.g, to a moving weather front); do lakes in different regions respond similarly to this specific episodic forcing event.

How long does the response persist? The way it is mediated is likely to depend on the trophic state of the lake (e.g., wind induced mixing and the consequent effects on the upwelling of nutrients and the phytoplankton production); and watershed characteristics (soil type and vegetation affecting e.g. the leaching of nutrients and the transport of particulate material).

Variables, that should at least to be monitored:

- good meteorological data

- good thermal data

- some measure of algal groups (e.g., bulk biomass; chl)

- algal production (estimation based on DO or CO2 measurements)

Group 3:

Complexity of interdisciplinary models (non-hydrostatic) and the difficulties of calibrating these models, particularly with respect to ecological parameters in physical-biological models.

Science question examples	What is lake metabolism?	What is effect of seasonal stratification on biogeochemistry	Propagation of an inflow through narrow reservoir	Biogeochemical effects of sediment resuspension in shallow lakes	1)How do surface and internal waves interact 2) What is wave-current interactions	Biogeochemical effects of sediment resuspension
Type of model (spatial)	Box model	One-dimensional (!D) water quality (WQ) model	2D (WQ) model (laterally averaged)	2D WQ model (vertically averaged)	3D model	3D WQ model
Bathmetry	Lake depth	Hypsographic profile	Fully grid-based bathymetry	Fully grid-based bathymetry	Fully grid-based bathymetry	Fully grid-based bathymetry
Forcing data - met data	Surface water temperature, DO	Daily shortwave and longwave (cloud cover) radiation, wind speed, rainfall, relative humidity - 1 station	Hourly shortwave and longwave (cloud cover) radiation, wind speed, rainfall, relative humidity - 2 stations	Hourly shortwave and longwave (cloud cover) radiation, wind speed, rainfall, relative humidity - 2-3 stations?	Hourly shortwave and longwave (cloud cover) radiation, wind speed, rainfall, relative humidity - 2-3 stations?	Hourly shortwave and longwave (cloud cover) radiation, wind speed, rainfall, relative humidity - 2-3 stations?
Inflow data	-	Daily (average) flow volumes, temperature, concentrations of simulated variables	Event scale (< 1 day) flow volumes, temperature, concentrations of simulated variables	Event scale (< 1 day) flow volumes, temperature, concentrations of simulated variables	Event scale (< 1 day) flow volumes, temperature, concentrations of simulated variables	Event scale (< 1 day) flow volumes, temperature, concentrations of simulated variables
Outflow	-	Daily (average) flow volumes	Event scale (< 1 day) flow volumes	Event scale (< 1 day) flow volumes	Event scale (< 1 day) flow volumes	Event scale (< 1 day) flow volumes
Calibration/validation data	Open-ended	Event based sampling or routine sampling or daily (averaged) profile data	2-D velocity data, temperature data (minutes), spatially varying simulated variables (hours to days)	Velocity data, temperature data (minutes), spatially varying simulated variables (hours to days)		Velocity data (seconds), temperature data (minutes), spatially varying simulated variables (hours to days)
Implementation in web system	Y

Morning session 8/3/05

IT Perspective: Frank Vernon, Barbara Benson, Fang Pang Lin, Ian Atkinson

Easily deployable systems

Smart sensors

Lack of spatial/temporal resolution of biological data compared to scales of processes.

Applications of omni directional camera

Remote downloading of full band width data

Stability of data communication under weather influence

Underwater wireless

Ability to reconfigure sensors in real time

Assessment of instrumentation status

Automated quality assessment/quality control

Adaptive sampling (what are the needs for this)

Adjustable data schema to automatically accommodate measurements changes

Bringing instruments to the grid using web services

Standards for interoperability (need to think through every level of the data to ensure transferability). "We need to eliminate arbitrary heterogeneity" [Barbara Benson]. Excel is an end point. Do we commit to making the data available in a web format.

- metadata

Policy requiring resource sharing (data, services)

Event detection: ability to issue warnings and to run models from real time data for forecasting purposes.

Need for personnel with truly inter-disciplinary skills (modeler, ecologists, IT)

IT development plans

- use case scenarios to derive requirements

- actual plans including priorities over various time windows

- identification of core technologies and services

- adequate resources (physical, software, people).

A question arising: what could we expect to get/achieve in one year?

Discussion

- Frank discussed the need to synchronicity across time for data logging devices.

- Smart sensors. Is the serial output system fundamentally flawed anyway?

- What types of data loggers?

- Need for discussion of open access to data, i.e. data format, ant an early stage

- Culture of data sharing. Consensus that value is in post-processing.

	Elementary (c. $5,000)	Intermediate (c. $30,000)	Advanced ($250,000?)
Budget, maintenance			Frequent sensor replacement
Science question - what can be done?	Seasonal dynamics Mostly non-responsive mode	Lake metabolism May be responsive
Hydrodynamics	Thermistor chain Onset Meteorological Station Inflow sensor	Campbell meteorological station Cabled thermistor chain Water level Net longwave Shortwave	· Wave gauge · High resolution thermistor chain · Current sensor (ADCP) · Comprehensive meteorological station · Water level · Conductivity sensor · OBS/ABS for turbidity,conductivity · LSST profiler adapted to in situ deployment · Inflow gauging · In situ flow cytometry
Chemistry		Dissolved oxygen, pH	· Dissolved oxygen, pH, colour, Automated nutrients (wet chemistry) · Nitrate/CO₂ probes
Biology		Single band fluorometer	· Fluorescence (single wavelength, multiple wavelength, PAM, FRR) · Satellite imagery · Video/Soar

Tuesday afternoon: Data integration and bringing this all together

Project Timelines

Who involved, from where, when?

Project Philosophy

Make data available to sciences as quickly as possible and in a sustained manner.

Data: Birth to Uses

Data generation transport, storage curation, databasing etc.

One simply administrative domain

Buoy(s), two-way wireless, antenna, internet filtering (QA/QC), database of sensor data loaded automatically, extend to other lakes etc.

Architectural design considerations

- autonomous administrative domains

- core data available in near real-time from sensors

- extension to new lakes

- standardize where possible

- automatic update of new data

Phase:

1) Establish a prototype using JDBC connections to low query of two lakes

2) We are redesigning the interface for two or more lakes to provide a registration of lakes into the larger system and to allow connections into the data in the database via web services. (March 2005)

Extensions:

1) Extending the system to other types of sensor data, e.g., ADCP data

2) Extension beyond sensor data to other ancillary data (e.g. lab chemical samples)

3) Instituting security features (data attack)

4) Integration with computational or presentation tools

5) Developing cross-site query tools

6) Linking data to other data (e.g. remote sensing data).

Sharing lake databases (Tuesday afternoon)

(Vlad Veyster and David Balsiger)

Would like to create a eon-stop portal to allow scientists to discover data from other lakes, query that data and apply computational models to it.

We hope that having such a comprehensive portal will improve collaboration and make for ready exchange.

Using database registration code. For our database registration we are using GridSphere portal framework 0 open-source network and standard compliant.'' (essentially JAVA nology).

Same query and data structure - variables etc. are the same for the two lakes.

Form:

(1) Database registration and put in web interface URL;

(2) List of tables that you are interested in - stores schema;

(3) Two data based options.

(4) Querying registered databases. Would like method to query registered data base. Show current search capabilities.

Do to lake metabolism data based.