Project Status

Task 0: Preliminaries

Task 0.a: Define problem

• Develop solid understanding of carbon cycle in N. Eurasia (processes, storages, fluxes, etc.) - Ted - DONE
  • For a summary, see Ted's Princeton Workshop talk
• Identify which processes/biomes etc are most important - Ted
  • Wetlands:
    • Most important carbon sink and methane source
    • A global model of wetland methane emissions as a function of climate does exist (Walter-Heimann)
  • Lakes:
    • A smaller contribution to methane production than wetlands
    • More complex behavior than wetlands:
      • They both produce carbon locally and import carbon from terrestrial systems
      • They both accumulate carbon in their sediments and vent carbon to the atmosphere
      • They produce methane, some of which is from recently fixed sources and some of which is from old sources
      • Phytoplankton form base of food web; phytoplankton behavior depends strongly on N and P concentrations
      • Fate of carbon depends on food web
    • Ted is currently investigating how to handle lake carbon cycling
      • Can we ignore them?
      • Are there models out there that can predict their behavior across N. Eurasia as a function of climate and other inputs that are available at large scales?

Task 0.b: Domain of interest

• NEESPI domain (approximate boundaries):
  • 15 E Longitude in the west
  • Pacific Coast in the east
  • 40 N Latitude in the south
  • Arctic Ocean coastal zone in the north
• Grid:
  • 25-km EASE grid polar projection
• Mask:
  • 25-km and 100-km EASE grid polar projection masks are available from Ted
• Regions of Interest:

  Region	Large Basin	Small Basin	Point Observations
  1. European Russia	Volga	Zapadnaya Dvina, Upper Volga	Central Forest Biosphere Reserve (straddles Z. Dvina and U. Volga, but intensive measurements, including Fyodorovskoe flux towers, fall in the Z. Dvina basin); Valdai water balance station (just north of U. Volga basin)
  2. W. Siberia	Ob	Chaya, Yeloguy/Dubces/Syum, Chulym	Bakchar Bog (Chaya), Zotino flux towers (Dubces), Hakasia flux towers (Chulym)
  3. E. Siberia	Lena, Kolyma, Indigirka	To be determined - possibly small basins examined in GAME studies at Yakutsk (larch forest) and Tiksi (tundra)	Yakutsk (Lena), Tiksi (Lena), Chokurdakh (Indigirka), Cherskii (Kolyma)

Task 1: Model Extensions

Task 1.a.1: VIC lake and wetland model improvements

• Modify VIC to handle unconstrained inundation - Ted - DONE
  • Testing of the lake model requires generating a depth-area relationship for inundation; this might be accomplished via satellite altimetry and satellite-derived open water classification for many years/seasons - see Task 3 for details
• Implement correct energy balance in soil beneath lakes - Laura, Krishna
• Add water table parameterization to VIC - Krishna, Laura
• Modify VIC to simulate transpiration from inundated vegetation (necessary for computing carbon cycling) - Ted/Laura?

Task 1.a.2: Model integration

• Insert BETHY fast ecosystem processes (photosynth, respiration, and phenology) into VIC - Ted
  • The plan is to add the appropriate functions and parameters to VIC. Photosynthesis is done and Ted is testing/comparing the results of Jarvis- and photosynthesis- based canopy resistance. Respiration and phenology are partially finished as well.
  • Ted is also working on a "baseline" of VIC and BETHY alone at the Fyodorovskoe and Zotino sites. Results from Fyodorovskoe suggest that further calibration of these models there is needed.
  • Reiner will help Ted set up a stand-alone version of JSBACH (not coupled to a climate model) for comparison to the VIC-BETHY combination. Ted will perform the JSBACH benchmark runs at the same set of flux towers as the VIC-BETHY tests.
• Send VIC/BETHY soil T, water table, and NPP estimates to the methane model - Ted
  • Ted has done this for the Valdai and Fyodorovskoe sites using all models (VIC, BETHY, and Walter-Heimann) on a stand-alone basis.
  • When the VIC/BETHY integration is complete, Ted will begin merging the Walter-Heimann model into VIC as well. This will allow a more sophisticated formulation of the Ro parameter (see task 1.b.1).
• Investigate whether we need to model the following processes, or if they can be approximated by simple assumptions - Ted
  • leaching of dissolved organic carbon from terrestrial soils
  • lake carbon cycling
  • exposure of fresh carbon in regions of permafrost thaw
• Integrate LPJ dynamic vegetation - Ted
  • Requires merging of characteristics of VIC veg types and LPJ pfts
  • Send VIC hydrologic state and BETHY plant growth data to LPJ at the end of each year (month?); let LPJ update pft fractions accordingly
  • This piece is not needed right away; for calibration/validation of present-day fluxes we can prescribe current vegetation conditions. We will need this piece for retrospective and future estimates.

Task 1.b.1: Methane model extensions

• Derive general function to predict Ro parameter - Ted
  • This function might contain explicit dependence on soil carbon pool content, as modeled in VIC/BETHY

Task 2: Data Preparation and Analysis

Task 2.a: In-situ Data

• Historical landcover summaries - Ted, Nina, Daniil
  • Region 1:
    • Have:
      • Upper Volga (Volga-1) in the Tver region
      • 28m DEM, Forest cover, peatland extent and soil textures from CFR
    • Need: Potentially all remaining Volga-1 regions
  • Region 2, 3:
    • SHI summaries not available; however we may be able to create a time series in these regions by using various other data sets, for example Larry Smith's peatland database (region 2), and lake extents derived from LANDSAT mosaicsfrom the early 1970s and 1990s in both regions.
• Local meteorological and hydrological observations- Ted, Nina, Daniil
  • Region 1:
    • Have:
      • Valdai water balance station 1966-1983 (include SoilM, WtrTbl, Runoff, Evap, SoilT)
      • Valdaj NCDC weather station 1949-1965
      • Fyodorovskoe TCOS flux towers 1998-2005 (include Evap, Soil T, CO2)
      • CFR ("Fedorovskoe") weather station 1963-2006
      • Sychevka (Evap)
      • Torzhok (Soil M/T)
      • Discharge from U. Volga at Eltsy
      • Discharge from Obsha (Z. Dvina basin) at Beliy
    • Need:
      • Discharge from the reservoir (which we must pay for)
      • CH4 emissions from the region (Ted is working on getting these, with Juliya's help)
  • Region 2:
    • Have:
      • Zotino TCOS flux towers 1998-2003 (include Evap, Soil T, CO2)
      • Vorogovo NCDC weather station 1982-2006
      • Discharge from Yeloguy at Yeloguy
      • Discharge from Syum at Syum
      • Bakchar NCDC weather station 1973-2006
      • Discharge from Chaya at Podgoenoye
      • Hakasia TCOS flux towers 2002-2004 (include Evap, Soil T, CO2)
      • Sira NCDC weather station 1969-2001
      • Discharge from Chulym at Balakhta
    • Need:
      • soil moisture/temperature from stations in this region
      • CH4 emissions at Bakchar Bog in Chaya Basin (Ted is working on getting these)
  • Region 3:
    • Have:
      • Yakutsk TCOS flux towers 2002-2005 (include Evap, Soil T, CO2)
      • Yakutsk/Jakutsk/Yakutsk_Magan NCDC weather stations
      • Yakutsk AAN/IORGC weather station (include Soil M,T)
      • Cherskii TCOS flux towers 2002 (include Evap, Soil T, CO2)
      • Cherksij NCDC weather station
      • Chokurdakh TCOS flux towers 2003-2005 (include Evap, Soil T, CO2)
      • Chokurdah NCDC weather station
      • Tiksi NCDC weather station
      • Tiksi AAN/IORGC weather station (include Soil M,T)
    • Need:
      • Identification of small basins used in GAME experiments & their discharge
      • CH4 emissions near Yakutsk (Ted is working on getting these)
      • CH4 emissions near Tiksi (Ted is working on getting these)
      • CH4 emissions near Cherskii (Ted is working on getting these)
  • Ted is searching for observational datasets (not necessarily from the NEESPI region) containing meteorology, water table, soil temperature, CO2 flux, and methane flux. This will allow better calibration/validation of the modeling framework.
    • Have:
      • Marcell Experimental Forest, Minnesota 1988-1990 (CH4, Soil T, WtrTable, Met data, Runoff, Evap)
      • George Reserve, Wisconsin (CH4, CO2, Soil T, WtrTable, Met data)
      • Fairbanks, Alaska (CH4, met data)
    • Need:
      • Salmisuo bog, Finland (CH4, Soil T, WtrTable)
      • Fairbanks, Alaska (WtrTable, Soil T)

Task 2.b: Remote Sensing Data

• Classification of JERS imagery - Erika, Ronny, Kyle, Daniil, Ted
  • Region 1:
    • Erika has produced an ascii file containing landcover classification for individual JERS pixels and sent it to Laura (Purdue) and Ted (UW). At first glance, the file's format looks good. Erika is also working on a file listing the acquisition dates of the pixels; however this is more difficult since ISPRA did not provide these for all pixels.
    • Validation of this calibration requires coordination with observations from CFR - Kyle, Daniil
      • JPL will provide landcover classifications to MSU
      • MSU will provide ground observations from the Central Forest Reserve (CFR) to JPL
    • The JERS images unfortunately have a large data gap running through the Upper Volga region. However, PALSAR data may fill this gap.
  • Region 2:
    • Erika has produced a landcover classification for the upper Ob region.
    • Kyle has found a landcover classification for the Zotino area, which could be used as a validation of the JERS classifications. However, the person at MPI Jena who made the classification posted on the TCOS web site is no longer in science, but a colleague will send some of the original data files to Kyle.
  • Region 3:
    • JERS data may be available here.
• Creation of multi-image history - Kyle, Erika, Ronny, Ted, Laura
  • PALSAR
    • Kyle and Erika have obtained 50 scenes so far, falling in several of the focus basins and acquired in summer 2006. We can obtain another 50 scenes at the end of March, with the goal of winter scenes so that we get an idea of seasonality.
  • LANDSAT
    • The JPL cartography lab has produced a mosaic of LANDSAT imagery from the mid-1990's which may be able to help validate the JERS classifications. The different resolution will require some resampling.
    • Ted has contacted Larry Smith (Region 2) and Katey Walter (Region 3) about using their LANDSAT mosaics acquired during the early 1970s.
  • MODIS
  • AVHRR
    • Ronny is working on wetland/open water classification of AVHRR images

Task 2.c: Other Datasets

• Meteorological forcings over the NEESPI region
  • Jenny Adam (UW) is working on generating these right now. The forcings go from 1950-2000.

Task 3: Model Testing and Evaluation

Task 3.a.: VIC Lake and Wetland Model Extensions

• Unconstrained lake expansion - Ted
  • An inundation depth-area relationship needs to be generated to run the model. Ted is looking into the feasibility of using satellite altimetry to get a depth timeseries. Meanwhile, hopefully PALSAR data can supply the time series of inundation extent.

Task 4: Retrospective Carbon Balance

Task 5: 21st Century Scenario Analysis

Timeline