management imperative. The global distribution of human impacts over time must be known. Key datasets include the following:
Land cover. The type and characteristics of global managed and natural vegetation must be known for the present, for the past (as best as can be determined), and from now forward (notably through monitoring, using high-resolution remote sensing). An historical record from the beginning of the industrial era is needed to assess the effects of land-use change on climate and the carbon cycle.
Land use. The characteristics of land management, including agricultural practices, forest management practices, irrigation, urbanization, and other direct human manipulations of ecosystems need to be known (historically, to the extent possible) and monitored. This information is important for modeling ecosystems, trace gas emissions, and water quality.
Fisheries' practices. It is becoming increasingly evident that fishing dramatically affects marine ecosystems and may alter their interactions with the marine physical environment. Marine ecosystem management practices must be documented as part of global marine ecosystem studies.
Atmospheric deposition. Atmospheric deposition has recently emerged as a significant issue in global change research and for over a decade has been viewed as an important ecosystem stress at regional scales. Deposition of important anthropogenic chemical species, especially sulfur species, nitrogen species, and ozone, must be documented.
The above types of data, describing the impacts of humans on ecosystems in a geographic framework, form a central part of the information needed for large-scale investigations in marine and terrestrial ecosystems. They complement other datasets, such as those on climate, soil properties, ocean circulation patterns, and natural vegetation-type distributions that are already widely used in modeling.88 Retrospective and contemporary data on human alterations and use of ecosystems and on collateral impacts are crucial not only for modeling the ecosystems affected but also because these data provide a critical link between ecological and human dimensions research (see Chapter 7).
Time series observations of atmospheric CO2 have long been a mainstay of global change research, and, as that time series has lengthened, the degree of insight it provides into the dynamics of the global carbon cycle has continued to increase. Time series measurements of other gases such as methane, nitrous oxide, and carbon monoxide and their isotopic composition are also crucial and have provided insights into their biogeochemistry.89