Records of fluctuating biotic and chemical characteristics from numerous sedimentary archives—including oceans (Guilderson et al., 1994), continental lakes, bogs, swamps (Hooghiemstra and van der Hammen, 1998; Behling and Hoogiemstra, 2000; Haberle and Maslin, 1999), and ice caps (Thompson et al., 1995)—demonstrate that tropical environments are dynamic at a range of timescales. One of the most recent extremes of these variations in geological time concerns the 100,000-yr glacial-interglacial cycles associated with the Quaternary geological period over the past 2.2 million years. Climate change involves massive reorganization of global climate systems with major impacts on ecosystem form and function that give rise to complex interactions between the atmosphere, geosphere, hydrosphere, and biosphere (Kohfeld and Harrison, 2000). Such changes can be documented by accessing sedimentary archives, whereas the nature and implications of such changes can be investigated through modeling. Numerous modeling approaches address different components of the Earth system. The majority of these initiatives have been developed within the temperate regions, particularly North America and Europe, with relatively few studies focused on the tropics. Indeed, as we will see, the modeled environmental history of the tropics remains poorly resolved despite its increasing importance in understanding global climates (Marchant and Hooghiemstra, 2004), biogeochemical cycles (Prentice et al., 1996), developing biogeographical theory (Tuomisto and Ruokolainen, 1997), and understanding issues concerned with biodiversity and human-environment interactions (Marchant et al., 2004b). Modeling climate change impacts in tropical environments represents a special challenge for computer climate models, in part due to the sharp relief of the mountain areas, where a single 1° grid-cell may encompass a wide range of environmental and climate gradients. These difficulties are compounded by the poor availability of marine and terrestrial data important for accurate simulations (Valdes, 2000). Nevertheless, as the complexity of the tropical
environmental system, and its importance to contributing to global climate models (GCMs), is being realized, models take an ever increasing number of parameters into account (Figure 6.1). However, such complex models require significant computing power and can be relatively static in time (non-dynamic). A compromise is available in models of intermediate complexity that can simulate climate evolution over thousands of years but use a relatively large grid-cell or a reduced number of inputs to attain a concomitant reduction in computational time.
This chapter will provide an overview of these different approaches, review the present understanding, and identify future areas for research development by investigating two main areas of modeling: first, biosphere models focusing on vegetation change, and the links from these to biogeochemical fluxes; second, climate modeling. First, we need to investigate the variety of modeling approaches—understanding model limitation is critical for future development and useful application (Peng, 2000). Such application may concern differences in the timing, intensity, and duration of the seasons that can have huge impacts on human prosperity, health, and surrounding environment. For example, the winter of 1982/1983 was exceptional: the dry seasons in Peru and Chile were very wet and the rainy season in Indonesia was extremely dry. Modeling can be focused to predict when a rainy season might fail, or when flooding or temperature extremes might be likely. However, if we cannot predict the weather next week, why should we trust climate predictions for next season, or in 25 years' time, or indeed 21,000 years ago! We can however say something useful about the climate trends, how certain components of Earth's system respond to such trends, and how these trends can influence our future climate associated environment and resources.
There are two basic modeling approaches: inverse and forward modeling techniques. In the former, comparisons rely on establishing empirical relations between modern and past environment observations through a transfer function (Kohfeld and Harrison, 2000)—that is, geological data are translated into climatic (e.g., mean monthly temperature, precipitation) or bioclimatic (plant distribution) parameters that can then be compared with simulated results. In the forward modeling approach, models are used to produce a predicted response that can be compared with current observation. These two approaches should not be seen as independent but as complementary (Kageyama, 2001); such a hierarchy of models is useful to allow the move from large to regional-scale investigation. This chapter will focus on modeling biosphere changes, although it should be noted that other components of Earth's system are required to understand these fully. For example, geophysiological modeling can assess the feedback mechanisms acting between geosphere and biosphere, the nature of the land surface, and impacts of changes in this—such as altering climate by albedo change that influences the reflectivity and absorption of incoming solar radiation, water vapor, convective precipitation, soil moisture, transpiration, hydrology, and latent heat flux (Figure 6.2).
Was this article helpful?