Tuesday Evening, 14 Sep 2004 – disciplinary working groups

I. Atmosphere group

A. What do we want out of this, what do we want from other people, what do we want to accomplish?

1. Construct carbon budget(s) on a range of spatial and temporal scales, and critically evaluate each component.

2. Atmospheric Transport

3. Flux components: photosynthesis, autotrophic and heterotrophic respiration

4. Optimize long-term sampling strategy

B. Backbone observing network should eventually be stand-alone

1. Develop and demonstrate methodology

C. Issues/questions for bottom-up team:

1. Will ecosystem fluxes be available at desired resolution?

a) hourly, 1-10km for regional studies

b) field-by-field for spatial scaling studies

D. HAVE (2005-2006)

1. Backbone

a) aircraft and tall towers, continuous measurements and flask samples

2. Aircraft Intensive Experiments

a) Flux mmts (sky arrow)

3. Crop-based C models

4. 20km improved transport

5. High Resolution Fossil Fuel model, EPA Air quality sites

E. HAVE NOT (2006)

1. Aircraft Intensive Experiments

a) Lagrangian experiments

b) Multi-species

2. Well-Calibrated Concentration measurements at some of the flux towers

3. Time resolved /temporally explicit/ bottom up crop models (also non-forest/non-crop)

4. Cloud-resolving transport (1km) for intensives

5. Validation of high resolution FF, Air quality experts

6. 5-Day Footprints: June 2000 Meteorology

7. All NACP Towers

II. Fluxes group

A. Goals:

1. Quantify and understand C stocks and fluxes of mid-continent

2. Participate in the development and testing of upscaling and downscaling methods

3. Quantify uncertainty in estimates of stocks and fluxes

B. What is needed to scale up and quantify uncertainty?

1. High precision/accuracy CO2 at tower sites for bl budget, CO2 profiles at towers,) bl dynamics (e.g. sodar)

2. Tier III (intermediate) data to help parameterize upscaling (LAI, foliar N, soil moisture, total biomass)

3. Reconcile flux and biometric estimates of NEP

4. Develop defensible means for quantifying uncertainty in NEP at multiple time scales

5. Modelers should work with flux people in development and testing of models (process, statistical)

6. Partitioning of fluxes using a variety of methods e.g. isotope measurements might be useful

7. Organization and structure needed for coordinated research

III. Data synthesis and modeling group

Tom Boden, discussion leader Norman Bliss, reporter

Round table on data synthesis and modeling needs for the Midwest intensive:

A. Verify atmospheric measurements with surface measurements

B. Eddy flux or canopy or satellite GPP or NPP

C. Statements on the degree of uncertainty

D. The draft NACP implementation plan includes goals for the intensives, but also want

1. more prediction: don’t limit to summer fluxes

2. Interdisciplinary groups: what time scale? daily, monthly, annual?

3. Eventually want annual and long term change. Then soil C is crucial.

4. The C budget problem of Pieter Tans is a multi-year problem; it can’t be addressed in a 2 year program. How does this Midwest intensive contribute to a multi-year view of the carbon budget for North America, in which the carbon history of the soils is important?

E. Examples of using existing data for carbon accounting or modeling:

1. Use NRI for crop rotation probabilities

2. FIA for forest biomass change

3. GOES thermal and shortwave: validated hourly insolation at 20 km

F. Data to share (currently being done):

1. Land cover change (carbon trends from land cover trends)

2. Satellite data to track land surface change (NDVI, etc.)

3. Daily maps of surface flux at 10 km resolution from thermal remote sensing and vegetation cover: disaggregate to tower footprint

4. For Ameriflux sites: GIS cutouts of STATSGO soil data, elevation, LAI time series, meteorological data, carbon flux, radiation, soil respiration

5. Fossil fuel emissions (state level time series monthly)

6. High resolution land cover maps (NASS: distinguish crop types)

7. Root zone available water capacity of soils or other soil properties.

8. For areas where SSURGO is not yet available, we know the distribution of soil map units and soil properties at a county level, but not the locations. We know the acreage counts of the arable land.

9. NASS data can be used in a protected environment [to ensure privacy], and then aggregated up to the desired spatial units.

G. Important data synthesis concepts:

1. Timing within the growing season influences fluxes

2. Seasonality of crop growth is tied to nutrients and water, and thus soils

3. Thermal inputs are useful for detecting vegetation stress

4. We want accurate estimate of C balances: a full budget

a) Crop yield that is being exported (grain)

b) Carbon exported in streams

c) N mineralization

d) Imports from fossil fuels, etc.

5. The large atmospheric fluxes drive the observation

a) Carbon modeling of the earth and the country

b) Need fast pools, residue amounts and soil climate in vicinity of the residue,

c) what affects plant growth

d) Not just data management problem, but a key scientific challenge to do the data integration

6. Reporting units (e.g., a county) may not be the same as the functional units (e.g., a flux tower footprint)

7. We want to use so many data sets that were not designed for the carbon problem: how can we use the data in a way that is true to the original statistical intent?

8. The soil survey program was not designed as a monitoring program.

9. It is crucial to assess or propagate uncertainties through model calculations and data processing.

a) This is not just representing data statistically, but includes the provenance of the data (trace to original measurements).

10. Spatial data sets may have inherently fuzzy boundaries.

a) A point-by-point (direct spatial overlay) may not be the best way to comnbie data sets with inherently different resolutions.

b) Look at atmospheric transport at the relevant time scales

11. Data and synthesis needs (important to do):

a) Change in species (e.g., agricultural crops)

b) Change in fertilization rates

c) Can use climate data (reanalysis at 6 hour intervals) to aid spatial scaling of flux towers?

d) Does this reflect processes at the surface?

e) Determine spatial scale for GIS layers (e.g., 60 m versus 1 km versus county).

f) Are land use history, soil, and climate data needed at the field scale?

g) Leaf area across the country

h) Standardize protocols and definitions for Ameriflux data

i) Emissions data (fossil fuels) at finer resolution (than state)

j) Legacy effects of long term land management activity

k) Include Federal lands in NRI

l) Resolve rangeland and forest definitions between NRI and FIA

m) Hyperspectral imagery (e.g., Hyperion) to distinguish tillage and residue types

n) Good soil data (accurate georeferencing)

o) Soil Data (SSURGO and pedon) in a format for national applications

p) Pedon data (real measurements) need to be linked to the map data in STATSGO and SSURGO (estimated properties)

q) Use existing data sets on soil, land use, elevation (and derivatives such as slope and aspect), climate, and related spatial data sets to help design the sampling scheme for intermediate (tier 3) sites. We should consider sampling based on the distribution of carbon stocks and fluxes, rather than the distribution of land area.

r) Consider inorganic carbon (carbonates) only if someone demonstrates it is important at less than geologic time scales.

H. Good discussion questions:

1. Can we measure photosynthesis from satellites?

IV. Uncultivated ecosystems group

A. Additional data & resource needs

1. Shared vision (defined experiment)

a) Coordination b/n atmospheric and terrestrial scientists

b) Defined:

(1) Scope

(2) Timeframe

(3) Space

(4) Inter- and intra-annual variation

c) Provide best estimate of carbon flux

2. Gaps in the program

a) “Uncultivated” is beyond forests

b) How important are these relatively infrequent systems?

3. Scaling

a) Temporal

b) Spatial

4. Other expertise (e.g., native grasslands)

a) Unclaimed systems

B. How data and gap-filling activities should be coordinated?

1. Broad agreement: effort will only be successful if coordinated.

2. Published methods (BioScience 2004)

3. Multiple approaches

a) Remote sensing

b) Towers

C. Does present agency funding process allow for coordinated research

D. Needs in Common (Identified by both Atmosphere and Fluxes groups originally)

1. High precision CO₂ measurements at flux sites

2. Need for time-resolved / spatially explicit bottom up models

3. Stable isotope measurements to partition photosynthesis and respiration

4. Temporal coordination of efforts across groups

5. NEE as the “interface” between bottom-up and top-down

E. High priority should be placed on coordination among participating flux sites

F. Common suites of measurements should be made at all sites

1. Timing, frequency, data availability

G. Coordinated modeling efforts at all sites

1. Suite of models, ID short list of key parameters to be measured

H. Process understanding - Below ground processes are still problematic and need to be addressed

1. Soil respiration measurements through year at flux sites

2. Separation of Ra and Rh via stable isotopes, trenching…

I. Consensus for a broad view of region

V. Cultivated ecosystems group

A. Restatement of meeting goal

1. Bring the atmospheric guys down to earth.

2. What important issues do we think the atmospheric folks might be missing?

3. Are we in a position to meet goals without more specific information?

4. What are the potential data sources? What new data are needed? What are our limitations/constraints? How do we put this all together?

B. Main Issues Discussed (in no particular order)

1. Spatial and temporal scaling

a) How do you scale from belowground (stocks, processes, dynamics) to atmosphere both spatially and temporally?

b) Our scale is largely at the process level. A key question is what happens to the functional C pools when you implement a management practice?

c) Atmospheric folks talking about a mixed signal. Difficult part may be separating the signal due to agricultural management from other signals.

d) Size of the area under consideration affects the sampling design. Need to note tradeoffs between various designs.

e) We can stratify land cover types by remote sensing. But still need observational data to characterize each type.

f) Ecosystem scientists can speak in terms of changes at an annual scale, but if top down guys need information on more temporally dynamic changes (e.g. responses to a cold front, short term events, etc.) that is more difficult for ground based measurements on a large scale.

2. Belowground vs. ground surface measurements

a) Needs to be some recognition of belowground processes as well as those that occur from ground surface up. Carbon gets stored belowground. Need to understand factors controlling storage and release and incorporate this into program.

3. Study boundaries

a) Use an ecological rather than a political boundary; basically, we are talking about the Corn Belt

b) Potential use of agro-ecozone maps by Sharon Waltman

c) Study areas should be constrained by footprint of the atmospheric tower.

d) This is a question for interdisciplinary group – need more from the top down modelers on what their constraints and most effective designs are – we don’t really have the same constraints based on lack of data.

4. Short-term vs. long-term fluxes

a) Short-term dynamics have more immediate effects on fluxes. Is the goal here to look at short-term (i.e., the length of an atmospheric campaign) or long-term patterns of storage or loss?

b) Do atmospheric folks want short term in order to match up with top down estimates?

c) If the objective is to tie into continental estimates, rapid cycling pools are more critical. However, long term pools are still important in terms of connections and driving forces.

d) Need to separate fossil carbon cycles (2 million yrs) from biological carbon cycle – have an opportunity to do management practices to impact biological. Focus on aspects of the carbon cycle that we have some control on.

e) How long is the campaign? Due to the potential for legacy effects from previous land use, previous residue inputs, and differing weather conditions, multiple year campaigns will probably be needed.

f) Over longer time periods, the potential for residues to move into pools that can ultimately be protected and sequestered can occur and this must be accounted for. But for short-term scales (1-2 years or less), rapidly cycling pools are likely to be more important for matching with top down approaches. Movement into slower cycling pools and stock changes are difficult to detect over short time periods (maybe need to wait 5 years or more to detect changes).

g) Ecosystem scientists can speak in terms of changes at an annual scale, but if top down guys need information on more temporally dynamic changes (e.g. responses to a cold front, short term events, etc.) that is more difficult for ground based measurements on a large scale. This is the same bullet that appears under scaling issues above but is relevant to both topics.

5. Variability

a) Many sources of variability in fluxes exist and operate at different scales. In addition to longer term (annual, seasonal) effects of climate, soil differences (e.g., drainage, texture), crop and overall management practices, specific short-term events (e.g., tillage and planting operations; major rainfall following a dry spell) can induce large variations in fluxes that may have effects well beyond the time scale of the event.

6. Other trace gases

a) Need to consider N2O. Two views:

(1) w/in context of North American Carbon Program, N was not included.

(2) However, in an agricultural system, everything is modified by N; plus N2O is arguably the most important greenhouse gas.

b) Need to consider the ways that N influences C, want to try to encourage nitrous oxide.

c) Tower people talking about measuring N2O

7. Stocks vs. processes

a) Annual fluxes – stocks vs. processes. Understanding processes is key to being able to account for different responses.

8. The role of models

a) We have models that will be tasked with doing the scaling up using ground based data and these models will give flux estimates, but instead of treating this as “truth”, need to recommend that there be more ground-based sampling in a systematic way throughout the campaign. If the purpose of this campaign is to set a precedent for future campaigns, then we should identify key locations that need to be sampled, who can do it and how it should be done. We don’t want to let models stand alone without additional measurements

b) How confident are we in the ability of models to handle effects of changes in climate and responses to management practices? – i.e., if we had intensive measurements, would it help the models? Need more known about spatial and temporal scaling.

c) It is possible to model short term dynamics by inferring from simulation/empirical models, but it is more difficult and we have less confidence in our estimates compared to longer-term dynamics (annual, decadal or longer).

d) Even if a model is perfect in predicting short term effects, a lot is dependent on high quality input data (e.g., small scale variability in distribution of rainfall inputs).

9. Sampling Details

a) How deep should we go?

(1) plow layer.

(2) Century normally goes to 20 cm.

(3) Root zone – need more information about deeper soils.

b) Inorganic carbon (effects of fertilizer application etc.) needs to be quantified in addition to organic carbon.

c) Bulk density is more difficult to measure accurately than carbon concentration and can contribute to large errors – need to address ways to reduce errors across study sites.

d) If we are doing more intensive sampling, how well does that have to mesh with the footprint of the tower?

e) SOPs for relating yield data to soils data to maximize value of the dataset.

f) SOP’s for sample depth increments – making sure that these match with other sectors

g) Would it be better to have several small towers to understand process, and then integrate smaller towers with large towers? Utility of aerial flights is that you can move aircraft faster than you can move a tower – such measurements might better capture short-term signals. Could aerial flights take advantage of signals induced by specific events such as tillage?

h) Need to make additional C stock measurements.

i) Combine remote sensing with some benchmark sites

j) What parts of the Corn Belt are managed the same? Chisel plow/tillage operation in April or May; heavy chiseling in Sept. This may have a large impact. Need to pin down timing of practices and then back up with remote sensing. If top-down is trying to pick up a signal in the atmosphere, then maybe timing atmospheric campaigns around these management practices would be useful.

k) Atmospheric folks are constrained – they want to understand vertical component – need to have a lot of data over several weeks. Can bottom up approach validate this? Bottom up would have to be made to coincide with their campaign or constructed such that they could be tied to other things that were measured at the same time.

l) Tall towers are continuous so it is not just entirely short term – bottom up, would this be expected to run for several years as well.

10. Potential Data Sources

a) Statistical aspects of sampling a field or identifying sampling areas (beyond simply soil series). What research groups might have data for their study areas or developed methodology that would help with sampling and scaling issues? e.g., Don Bullock out of Illinois, spatial stats; Univ. of Nebraska group working at Shashi Verma’s flux sites.

b) GraceNet sites – mostly small chamber, biomass, soil temperature, climate,

(1) Baker in St Paul

(2) Morris

(3) Ames

c) In addition to new GraceNet sites, there is still a role for more traditional long-term experiments.

d) Precision ag studies and implementation – provides detailed data on yield, soil measurements, etc. Provides an opportunity to tie in management system influences on productivity – there is a lot of process level information there. Precision ag committee. ARS group on precision ag.,

e) Keith Paustian and John Baker’s study in Iowa

11. Data Needs

a) Need to identify which sites are available for these studies. What existing sites could we tweak to fit? Look at where these sites are located spatially and then identify gaps (but at this point probably has to be without any additional funding).

b) Speak with Jeff Goebel – can we build out NRI to incorporate additional measurements (e.g,, manure applications). If they want to pilot techniques, this might be a good opportunity.

c) Benchmark soils – emphasize utility of having benchmark soils (similar to FIA) that can inform the models

d) Use NRI for land use; add soil measurements

e) Chicago Farm Bureau, leases – (Chuck Rice) – benchmark sites

f) Rich Conant, 2 papers on optimization of sampling at benchmark sites. Inverted triangle approach. In Environ. Management, JEQ

g) Land use history data – important. e.g., drainage tiles; past management practices etc., – gather knowledge from local resources

(1) Baker in St Paul

(2) Morris

(3) Ames

h) In addition to new GraceNet sites, there is still a role for more traditional long-term experiments.

i) Precision ag studies and implementation – provides detailed data on yield, soil measurements, etc. Provides an opportunity to tie in management system influences on productivity – there is a lot of process level information there. Precision ag committee. ARS group on precision ag.,

j) Keith Paustian and John Baker’s study in Iowa

12. Data Needs

a) Need to identify which sites are available for these studies. What existing sites could we tweak to fit? Look at where these sites are located spatially and then identify gaps (but at this point probably has to be without any additional funding).

b) Speak with Jeff Goebel – can we build out NRI to incorporate additional measurements (e.g,, manure applications). If they want to pilot techniques, this might be a good opportunity.

c) Benchmark soils – emphasize utility of having benchmark soils (similar to FIA) that can inform the models

d) Use NRI for land use; add soil measurements

e) Chicago Farm Bureau, leases – (Chuck Rice) – benchmark sites

f) Rich Conant, 2 papers on optimization of sampling at benchmark sites. Inverted triangle approach. In Environ. Management, JEQ

g) Land use history data – important. e.g., drainage tiles; past management practices etc., – gather knowledge from local resource

Wednesday Morning, 15 Sep 2004 – Interdisciplinary working groups

VI. Group 1

A. Remote sensing

1. Classification of 2004 remote sensing data, flip corn/soy, for planning-MODIS 250m data are fine as product

2. Clouds and delivery times could inhibit collection of remote sensing for intensive

3. Microwave data for soil moisture is important for aircraft/regional fluxes

B. Ground measurements

1. Need Synthesis of NASS/FIA/NRI data for region

2. Set up long-term measuring sites because need five years to see minimum changes

C. Simple statistical models for forecasting

1. Simple models can be used for planning aircraft flight measurements, photosynthesis, light saturation, respiration, temperature effect, soil moisture effects

2. Must be reasonable,

3. locate anomalies for investigation, aircraft data

D. Modelers challenged

1. Most model comparisons run after the fact based on experimental data – primarily because we need actual climate data

2. Based on previous data and experience with model comparisons, we can run model comparisons before experiment

3. Locate areas of divergence and test with field plots, tower locations, and aircraft flux data

4. Some modelers would rise to this challenge

5. Models would have algorithms improved

E. Gaps

1. Stream transport of carbon – identified in NACP PI 2003 meeting

2. Erosion-deposition needed more planning

VII. Group 2

A. What do we need from each other?

1. Focus on CO2 although there are also other trace gas fluxes of interest that may need to be considered

2. Top-down

a) Tall Towers – can integrate measures from daily to full time span of campaign

b) Aircraft – short time span needing more information on daily or even hourly

3. Bottom-up

a) Focusing on monthly, annual and longer time scales

b) Models exist that simulate carbon dynamics at finer time scales but activity data are often limited

c) Can assess uncertainties

B. Opportunity

1. Build on current bottom-up modeling as part of the NACP with some finer scale modeling for the intensive campaign

a) Activity data

b) Timing of tillage, manure applications, etc.

c) Finer scale data on weather conditions that are essential for capturing the daily fluxes due to dependence on moisture and weather

C. Collaboration

1. Bottom-up and top-down teams to integrate information, evaluate findings addressing consistencies and differences

a) Refine modeling of each group to better understand processes

(1) Rectifier effect

D. Dissolved organic carbon, Inorganic carbon, erosion and other uncertainties in the bottom-up modeling

VIII. Group 3

A. The overall approach ought to consist of multiple, complimentary investigations at several scales

1. Example from WI group: coordination of towers, flights, remote sensing (Martha Anderson)

2. Gain confidence based on footprint analysis, apply for model development

3. We’re lacking high resolution maps of LAI

B. Within each investigations data collection and format should be uniform across investigations to promote collaborative idea development and testing of ideas

1. Several different bottom-up approaches, several different ways of scaling up, various tower sizes,

2. Examples from other LBA or BOREAS?

3. Unique challenges for the Midwest Intensive

4. Coordination between data sets (e.g., if a ground-based study, flight, flux measurements require precise coordination)

5. Should tall towers be co-located w/ ‘regular’ flux towers, data collection at several levels (soil pits, whole canopy measurements, county yield)?

6. Multiple levels of sampling, how do we match these? This is the fundamental goal of the Midwest intensive (?)

7. Multiple models at each site – testing models, model integration

8. Benchmarks?

9. Scaling: w/ known land use within an area, run model for each component and sum to footprint

10. Unclaimed areas are a challenge

11. Need coordination efforts

12. Need management data in addition to land cover data

13. Will never have same quality data at broader scales

C. Multiple modeling approaches implemented for all investigations

1. Running 3-4 models at each site (Derek) of various spatial resolutions would inform each other (?)

2. Field model, site scale, Midwest scale, really large scale

3. Should be common set of models applied across the scales (requires close coordination)

D. How much data is currently available to answer these questions (Carl)

1. Flux sites

2. Land use, land use history, climate, soils

3. Missing data for the 40% of the region that isn’t cultivated

4. Urban fluxes? Urban forests?

5. We have NRI/FIA data, can collect additional data such as… (Kathy)

6. What satellite data are necessary?

a) LANDSAT would be great, but MODIS is the best bet, GOES for regional scale

E. Even if the data agree, how well does it reflect what happens in nature?

IX. Group 4

A. Need from each other? ISSUES:

1. agreement on units, terminology, time scales, spatial scales

2. common drivers and events driving fluxes

3. tracers to identify sources of fluxes forests

4. urban impacts/point sources

5. Get handle on messy/confounding areas (priorities)

6. Sub-urban

7. Focus on CO2 although there are also other trace gas fluxes of interest that may need to be considered

B. How good does data have to be?

1. Quantify uncertainty (across scales)

2. “Balance” precision (rank/stratify areas as re impact)

3. Optimize use of available/historic data

C. Standardized reporting?

1. YES!!

a) need data management system (on going activity)

b) land classification agreement of definitions

D. What kind of outputs will be produced?

1. YES to all mentioned plus:

a) Synthesis publications

b) Data archive driven by Information Quality Act

E. OTHER??

1. Effect of intensive water management (C & water quality)

F. Scaling

X. Group 5

A. Goals

1. Quantify and understand C stocks and fluxes of mid-continent

2. Participate in the development and testing of upscaling and downscaling methods

3. Quantify uncertainty in estimates of stocks and fluxes

Wednesday Morning, 15 Sep 2004 – Disciplinary working groups

XI. Atmosphere group

A. How can we turn "carbon cycle weather" into "carbon cycle climate"?

1. We are making measurements of short-term phenomena, such as the response of ecosystems to PAR, recent precipitation, drought stress etc., whereas what we want to know manifests itself on annual and longer timescales, such as the effect on soil carbon of management practices (present and past).

2. One example of this: for each separate year one can plot photosynthetic rate as measured by eddy covariance against PAR using only a few coefficients to characterize the relationship. The curve fit may slowly evolve over time in a statistically significant way. The latter could tell us something about the "carbon cycle climate".

3. It is of course crucial that the measurements have "climate quality" which is typically about 10 times more demanding than "weather quality". If not, we could be mistaking instrument drift or methodological changes for a real trend.

4. A second point pertinent to the opening question is that long-term drivers have to be incorporated into models that are compared to, or fitted to, the measurements dominated by short-term weather.

XII. Fluxes group

XIII. Data Synthesis and modeling group

A. Keys:

1. Need capability to explain

a) identify key processes that control fluxes

b) short- and mid-term scales

2. How to assimilate data into models

a) Initial condition

b) parameter est.

c) provides info. about land use [historical] legacy

d) model adjustment - pull data back into reality

3. Uncertainty analysis

B. Considered two basic types of models

1. Statistical Models

a) These do have a role

b) Not inaccurate and incorporate basic knowledge

c) Useful for flux tower [e.g., Wylie in rangelands]

2. Process models

a) crop/SVAT

b) biogeochemistry/agro-ecosystem

c) remote sensing based

C. GOES(thermal?) & LANSAT [U. Wisc., Martha]; Disagregation; multi-scale modeling approach; Net carbon flux; Link tower and aircraft data; Uses some empirical relationships, Light use efficiency concept/model with CO2 canopy gradient adjustment

D. Most biogeochemistry & land surface models have not been used for ag situations – what is their role?

E. Effects of tillage – erosion, residue/decomposition, timing of nutrients during growing season [need daily model, proper timing for availability]; allocation & phenology

F. What kinds of modeling activities would be appropriate for NACP Intensive

1. Before Intensive starts [pre-campaign]

a) Identify appropriate models through model data comparison activity

b) Perform data assimilation

c) tune up for agro-ecosystwms

d) what are key parameters that will need review, tuning? stratify models to determine data requirements

(1) for example, LAI [available through MODIS – not real reliable, so can/should “verify” during on-site visits]

2. Modeling in coordination with atmospheric data analysis

3. Modeling for long-term NACP Goals

a) Important to keep in mind longer term projections, including capabilities for analyzing effects of policy proposals

b) Scaling up

c) Keep in mind the uncertainties

d) Need to worry about the modeling up front rather than after the data collection; need a “Leader”

XIV. Uncultivated ecosystems group

A. How to deal with uncultivated and “messy” lands?
Wednesday Morning

1. Forests

2. Grasslands

a) Native

b) Pasture

3. CRP/WRP land (~15 M ha nationally)

4. Riparian

5. Developed land (golf courses, suburban)

a) Data for suburban lands from NRCS

b) Fertilizer inputs 10% of total

6. Agriforestry (wind breaks, wood lots, etc.)

a) Lack of data re. extent and flux

7. Maybe determine area by difference?

8. How to characterize these lands?

a) by % of study area

b) by % of total sink strength

B. Additional Points:

1. Potentially arrange bottom-up work along broad transects

a) Iowa-Minnesota

2. Other Processes we need to be aware of:

a) entrainment & deposition

b) land use legacies

XV. Cultivated ecosystems group

A. Priorities

1. Understand the spatial and temporal variability patterns in fluxes in cultivated systems.

2. Quantify the impacts of soil differences on fluxes (CO2, H20, and GHG’s)

3. Variation in land management practices and interaction with response need to be quantified as part of flux studies.

B. Soils in formation for processes at all scales

C. Anthropogenic decisions on land management

D. Enhance utility of microclimate data bases

E. We need to better understand “C-cycling” under a wide range of climatic scenarios-wet/cold, wet/hot, dry, and also simply better understanding many of the processes in the soil.