Mississippi Basin Carbon Project Science Plan

Land surface characterization and analysis

The available data described above will require substantial integration and analysis before application to the research objectives of the MBCP. Understanding the transport of water, sediment, and carbon across the land will require a thorough characterization of many aspects of the land surface. Before this characterization can be accomplished, the available spatial data must be closely scrutinized and evaluated for accuracy and internal consistency. Methods must be developed to adapt and integrate data that were collected and compiled for a wide range of purposes. For example, the assignment of watershed boundaries must take into account the delineation of USGS hydrologic unit codes (HUC’s), topographic information from DEM’s, and the river-reach files of the Environmental Protection Agency (EPA). Because the required characterization of the land surface will have utility for a wide range of purposes in addition to our own, we expect that our integration and analysis of land-surface data (utilizing GIS techniques) will be a very important byproduct of the MBCP.

Table 1: Spatial data available for the Mississippi Basin Carbon Project

Our analysis will require integration of spatial data with a diverse array of information. The complexity of this task can be illustrated by considering the kinds of data integration needed to examine the hypothesis that significant quantities of carbon are being deposited and stored as sediment on the landscape. These include:

• Lake and wetland sedimentation: The primary source of information for lake and wetland sedimentation is the North American Pollen Database maintained by National Geophysical Data Center in Boulder, Colorado. Presently, about 500 sites are included. Data can be interpreted to provide information about both land-use and sedimentation rates through time. The catchment basin corresponding to each pollen site can be characterized within a GIS and DEM. It should be noted that lakes are probably a minor player in the human-modified carbon cycle, whereas reservoirs may be much more significant (Mulholland and Elwood, 1982; Kempe, 1984; Stallard, in press).
• Reservoir sedimentation: Two databases must be integrated into a GIS and DEM for the Mississippi River Basin. These are the National Inventory of Dams (NID) (FEMA and ACE, 1993) and the Sediment Deposition in U.S. Reservoirs, Summary Data Reports of the Subcommittee on Sedimentation, Interagency Advisory Committee on Water Data. The second database is maintained by the NRCS and is now referred to as the Reservoir Sedimentation Survey Information System (RESIS) (Steffen, 1996). The NID database contains state-by-state information about all dams in the United States that are 1.83 m (6 feet) or greater in height. This data base must be standardized: units should be metric; all dams must be located using longitude- latitude coordinates; and all dams must be cross-referenced to identified reservoirs and watersheds. The RESIS database documents changes in reservoir storage capacity throughout the United States, and contains information gathered by disparate types of surveys. This is the master data base for regional erosion studies, tracking about 1,800 reservoirs. Information was interpreted for the period ending in 1965 by Dendy and others (1973). Recently, Renwick (1996) reexamined the database through 1975, and concluded, "Human impact appears to be much more significant in controlling spatial patterns of sediment yield than natural factors."
• Discharge: Many agencies compile composition data from surface waters within the Mississippi River Basin. We will compile selected discharge data from USGS sources. Our starting point will be the compilations by Alt (1994, 1995). Selected sites should have long term records and attention should be given to sampling during discharge events. Water Resources NASQAN and NAWQA sites will receive special attention, as will sites where data were gathered during major historic floods such as the 1993 floods. At each site we will try to establish a relation between total suspended solids and carbon concentration of the sediment. The more robust sediment record will then be used to interpolate estimates of suspended carbon loads for periods when carbon was not directly measured. Nutrient and alkalinity discharge will be compiled in a parallel manner. The scope of discharge data utilized in this manner will be expanded to include less robust data sets if necessary to better interpret results.
• Atmospheric nutrient loading: The National Acid Precipitation Network (NAPAP) has monitored atmospheric loadings of dissolved constituents in precipitation, including the important fertilizers, nitrate and ammonia. Presently, we do not know whether loading maps have been prepared or whether we need to prepare them. Mapped data can be compiled in a GIS so that loadings may be determined by watershed.
• Fertilizer application, other agricultural information, and demographic information from county-scale data bases: Fertilizer application, legume crops (soy), and domestic wastes are sources of fixed nitrogen and phosphorus, the primary ecosystem nutrients. Information about these sources is compiled on a county basis by the US Census and the Department of Agriculture. These data can be used to estimate nutrient loadings.
• Characterization of weathering sources: GIS-based maps of geology, precipitation, runoff, and mean annual temperature can be used to model weathering sources of nutrients and sediment.
• Sedimentation in the Lower Mississippi River valley and delta: In the lowland part of the Mississippi River valley (approximately from Cairo, Illinois to the ocean), sedimentation is controlled by sea level, weak channel gradients, and channel modification. As such, the styles of sedimentation of the lower Mississippi River valley and delta are fundamentally different from the upland sites. The map of Saucier and Snead (Saucier and Snead, 1989), depicting Holocene sedimentation in the lower Mississippi Valley, can be digitized. Hydrologic modifications (such as levees, cutoffs, and sediment ponds) have been summarized by the Army Corps of Engineers in a series of reports. Using information from the Army Corps of Engineers and other sources, hydrologic modifications and contemporary land use of the valley can be superimposed upon the digital map using GIS techniques.

Further data integration and analysis will be necessary to examine the hypothesis that eroded soil organic carbon is replaced by newly fixed organic carbon. For example, the data sets and analyses discussed above can be organized with GIS methods and used as inputs to models of the dynamics of soil organic carbon on the landscape. An initial estimate of the quantity of soil organic carbon can be made from the STATSGO data base (Bliss, et.al, 1995), augmented with forest litter data from the U.S. Forest Service. In this manner the soil organic carbon content at various depths can be estimated, reflecting the NRCS program of sampling and database building over the last 40 years. To develop a more dynamic view of soil organic carbon, the data can be structured for use in time-dependent models. Spatial data sets for the factors that influence inputs to the soil organic carbon pool; including type of vegetation, soil properties, climate, and management practices; can be interpreted and adapted using GIS techniques to estimate the spatial distribution of rates of organic carbon input for soil models. The same data sets can also be used to develop relationships for estimating model soil organic carbon decomposition rates. The data can be evaluated in additional ways to provide inputs to models that simulate runoff, erosion, and sediment transport and deposition. The understanding of key processes will be strengthened by field research, and both spatial and field data will be used to parameterize and calibrate a variety of models, as discussed in the following sections.

	Contents
	Available data
	Identification and field study of key processes