nisms, whether chemical, biological, or physical, that control the Earth system and its climate; defining the degree to which the Earth system has changed and is changing over periods of years, decades, centuries, and millennia; and (c) exploring largely uncharted regions, which may be defined geographically, mechanistically, or in other scientific terms. In the course of establishing an observational approach it is essential not to lose sight of this distinction.
There are also important distinctions in the required datasets for the different disciplines. These distinctions are the basis of fundamental “cultural” differences in the architectures selected for specific observational approaches. For example, observations are obtained in different ways to address questions about different phenomena, such as the following:
Ice cover changes as a function of time on seasonal-to-decadal scales.
Free radicals at the parts-per-trillion level in the troposphere and the stratosphere.
Vegetation pattern changes in terrestrial systems and oceanic systems.
Secular trends in atmospheric temperature with an accuracy of 0.1 K, as a function of altitude, latitude, longitude, and season.
Mesoscale meteorological events tied to global-scale variations such as the El Niño-Southern Oscillation (ENSO).
Observations required for each of these phenomena are not obtainable through a single solution, such as a single global network of ground-based observations or an ensemble of space-based remote sensors. While considerable intrinsic programmatic pressure exists for a “unified ” solution to Earth observations, the scientific context speaks strongly for a flexible and adaptive aggregate of techniques that attack specifics, whether of long-term trends or of mechanisms that control the Earth system.
A series of examples in Chapter 2, Chapter 3, Chapter 4, Chapter 5, Chapter 6 through Chapter 7 also represented a broad spectrum between observational constraint and theoretical speculation. The scientific method is pursued in an effective and vital manner when the fundamental design of the observational approach is matched to the calculated observables such that specific mechanisms, fundamental to the system, are tested directly. Models are very powerful when used in this context (see Chapter 10). They are central partners with observations in the course of proving or disproving fundamental assumptions.
This report approaches the problem of observations as a synthesis, working from scientific research needs to observational implementation. It has always been assumed that building a global observing system would serve the needs of most of the science components of the USGCRP. Indeed, a parallel activity is taking place (Global Climate Observing System, GCOS) to design a global observing system for climate to satisfy both scientific and monitoring needs.1 Parallel efforts are under way for the ocean (Global Ocean Observing System) and