The opinion, quasi-general, is that … the areas … north of Macenta, Gueckedou, Kissidougou will soon be no more than a vast poor savanna, the [forest] islands and gallery forests still present at risk of being rapidly destroyed.

The region of Kissidougou was covered by a deciduous forest of Khaya sp., Chlorophora sp., Antiaris africana, Afzelia africana, Ceiba pentandra, Triplochiton scleroxyllon …

Around 1945, the forest, according to the elders, reached a limit 30 km north of Kissidougou town. Today, its northern limit is found at the level of Gueckedou- Macenta, thus having retreated about 100 km … This deforestation is essentially the result of human action.

In the green belts which surround the villages, one finds the relics of original primary forests. The value of these biotypes in the heart of a nearly 100% degraded environment is inestimable. One finds no individual of [characteristic savanna tree species] more than 35 years old … supporting the thesis that the site has burned systematically only since then.

ABSTRACT One of the major problems in macroscale hydrological modelling is the assessment of areal heterogeneity in important land surface characteristics, such as topography, land use, land cover, soil, vegetation and hydrological characteristics. After a brief discussion of spatial scales to be covered and related categories of models to be applied two examples are presented which underline the problems involved in the application of lumped models for large areas, like grid areas of general atmospheric circulation models (GCMs). A strategy for a more appropriate hydrologically sound structuring of macroscale hydrological models is then outlined which takes into account the following facts and features: (1) zones of ‘uniform’ atmospheric forcing, (2) landscape patchiness, (3) intra-patch heterogeneity, (4) a ‘Two-Domains-Modelling’ concept, which is essential for the coupling of atmospheric and land-surface hydrological models.

ABSTRACT Partial analysis in the form of limiting analytical solutions is applied to the problem of infiltration into a semi-infinite column at constant initial moisture content. It is found as a result of such an analysis that: (a) the solutions for realistic assumptions concerning the soil moisture characteristics are bounded reasonably closely by an upper limit corresponding to constant diffusivity and ultimate infiltration rate equal to saturated conductivity and a lower limit corresponding to constant diffusivity and zero conductivity; (b) for both of the limiting cases of zero horizontal conductivity and infinite horizontal conductivity, the values of average sorptivity are insensitive to the form of the statistical distribution of spatial non-homogeneity; (c) the average sorptivity is less than the corresponding sorptivity based on the average scale parameter of a spatially variable soil and the average ultimate rate of infiltration is greater than the corresponding ultimate rate based on the average scale parameter.

ABSTRACT Hydrology has important contributions to make to the development of global models by providing independent calibration of land surface components of global circulation models (GCMs), by validation of GCM outputs at the basin level through comparison of recorded and simulated streamflows and by examining the implications of climatic change scenarios on water resources. Most importantly, hydrological models are necessary to provide the lateral links needed to close the land surface/boundary layer feedback loops. At grid scales, lateral water fluxes significantly affect soil moisture availability for evapotranspiration. At regional scales, accumulated runoff determines the freshwater inflows to the oceans which drives the sea-ice cover and ocean salinity flows. A hierarchy of hydrological models (HHM) has been developed using a ‘grouped response unit’ (GRU) approach to link process parameters to land cover, basin topography and the areal extent of climatological phenomena. The GRU allows land-use/land-cover to vary from element to element within a unit. Analyses of data suggest a grid element in the order of 10 km × 10 km is appropriate for hydrological modelling. Temporal resolution is accommodated by using three hydrological models operating at hourly, daily and monthly time scales.

ABSTRACT Mesoscale circulations generated by landscape discontinuities (i.e. ‘seabreeze’ like circulations) are likely to have a significant impact on the hydrological cycle, the climate, and the weather. However, these processes are not represented in large-scale atmospheric models (e.g. general circulation models (GCMs)), which have an inappropriate grid-scale resolution. Assuming that atmospheric variables can be separated into large-scale, mesoscale, and turbulent-scale, Avissar and Chen (1993) developed a set of prognostic equations applicable in GCMs for momentum, temperature, moisture, and any other gaseous or aerosol material, which include both mesoscale and turbulent fluxes. They suggested using the mean mesoscale kinetic energy (MKE) per unit of mass for parametrizing these mesoscale fluxes in such models. In the present study, which complements the work of Avissar and Chen (1993), we simulated the atmospheric planetary boundary layer (PBL) that develops above a locally deforested area of the Amazonian region, to illustrate the relationships that exist: (i) between the diurnal variation of MKE and the diurnal variation of mesoscale latent and sensible heat fluxes; and (ii) between landscape discontinuities resulting from horizontal gradients of moisture at the ground surface and MKE. We compared MKE with turbulence kinetic energy (TKE) to emphasize the magnitude of mesoscale processes, as compared to turbulent processes. This analysis illustrates the potential use of MKE to bridge between landscape discontinuities and mesoscale fluxes and, therefore, to parametrize mesoscale fluxes generated by such subgrid-scale landscape discontinuities in GCMs and other largescale atmospheric models. […]

ABSTRACT Flood plains are regions which are strongly dependent on the hydrological conditions induced by the dynamics of the fluvial water system. These regions also exhibit a broad spectrum of different habitats closely interfaced and of large biological diversity.

The objective of this paper is to describe some relationships between the dynamic characteristics of hydrological variables and the spatial variability in fluvial ecosystems. The interrelationships are not well understood but some hydrological key variables can be identified which drive the biological system. The integration of temporal hydrological characteristics with the spatial variability of some abiotic parameters, such as soil type, yields an indication for the spatial variability in the environment. The links between the hydrological system characterized by the soil moisture budget and the vegetation layers in the riverine forest ecosystem are analysed at the patch scale and the regional scale.

The methodology is applied to a flood plain region located along the Austrian section of the Danube. Due to the implementation of a hydropower scheme the backwater region was separated from the main river by impounding dams located along the old river banks. The temporal pattern of the hydrological variables, including the surface water table and the groundwater table, was completely altered. The changes in both the abiotic and biotic systems were monitored over a rather long time and some links between these systems were studied.

ABSTRACT Geophysical phenomena are often characterized by complex, random looking deviations of the relevant variables from their average values. Traditional approaches attribute this complexity to external uncontrollable factors and to poorly known parameters, whose presence tends to blur some fundamental underlying regularities. In this paper we consider that complexity might be an intrinsic property generated by the nonlinear character of the system's dynamics. Bifurcations, chaos and fractals, three important mechanisms leading to complex behaviour in nonlinear dynamical systems, are reviewed and the role of the theory of nonlinear dynamical systems as a major tool of interdisciplinary research in geosciences is stressed. The general ideas are illustrated on the dynamics of fluctuation-induced transitions between multiple climatic states with special emphasis on Quaternary glaciations.

ABSTRACT There are two primary approaches to modeling rainfall; stochastic modeling and deterministic integration of nonlinear partial differential equations which model the atmospheric dynamics. The statistical advantages of the former could be combined with the physical advantages of the latter by exploiting cascade models based on scale invariant symmetries respected by the equations. Carried to its logical conclusion, this approach involves considering the atmosphere as a space-time multifractal process admitting either a vector, tensor or even only a nonlinear representation. The process is then defined by two groups which respectively specify the rule required to change from one scale to another and the corresponding transforms of fields. Both groups are characterized by their generators, hence by their Lie algebra. We show how to extend existing cascades beyond scalar processes, showing preliminary numerical simulations and data analyses, as well as indicating how to characterize and classify the scale invariant interactions of fields.

ABSTRACT A new land surface hydrological parameterization for atmospheric General Circulation Models is introduced. The model incorporates physically-based relations for the partitioning of atmospheric energy and moisture forcing. Using statistical-dynamical derived-distribution techniques, closed-form and computationally-efficient expressions are developed for the inclusion of subgrid-scale spatial variability into parameterizations for atmospheric models. Numerical experiments with a General Circulation model show improved water balance estimates.

ABSTRACT The effects of small-scale heterogeneity in land surface characteristics on the large-scale fluxes of water and energy in the land-atmosphere system have become a central focus of many of the climatology research experiments. The acquisition of high resolution land surface data through remote sensing and intensive land-climatology field experiments (like HAPEX and FIFE) has provided data to investigate the interactions between micro scale land-atmosphere interactions and macroscale models. One essential research question is how to account for the small-scale heterogeneities and whether ‘effective’ parameters can be used in the macroscale models. To address this question of scaling, three modeling experiments were performed and are reviewed in this paper. The first is concerned with the aggregation of parameters and inputs for a terrestrial water and energy balance model. The second experiment analyzed the scaling behaviour of hydrological responses during rain events and between rain events. The third experiment compared the hydrological responses from distributed models with a lumped model that uses spatially constant inputs and parameters. The results show that the patterns of small scale variations can be represented statistically if the scale is larger than a representative elementary area scale, which appears to be about 2–3 times the correlation length of the process. For natural catchments this appears to be about 1–2 km2. The results concerning distributed versus lumped representations are more complicated. For conditions when the processes are non-linear, lumping results in biases; otherwise a one-dimensional model based on ‘equivalent’ parameters provides quite good results. Further research is needed to understand these conditions fully.

ABSTRACT This paper gives a general overview of how parameter optimization using inverse modelling can be used to infer scale-dependent effective soil hydraulic functions. In this approach, it is assumed that macroscopic flow behaviour as represented by Richards' equation is valid at all spatial scales. Thus, solution of the continuity equation with appropriate boundary conditions is independent of the size of the physical system. In this manner, for example, watershed effective soil hydraulic properties can be estimated. Since the optimization requires areal-averaged boundary conditions and measured areal-averaged flow attributes, such as water content, it is shown how remote sensing can be applied to infer these area-averaged hydrological variables.

In this way effective soil hydraulic properties are obtained from inverse modelling using the dynamical one-dimensional soil–water–vegetation model SWATRE. Areal evaporation can be obtained from reflection and thermal infrared remote sensing, while areal soil water content is estimated from microwave remote sensing techniques. The combined remote sensing and inverse modelling approach is illustrated with data taken from the HAPEXEFEDA experiment in Spain and from the catchment Hupselse Beek in the Netherlands. It is concluded that small-scale soil physics may adequately describe mesoscale behaviour.