Close Window

SH-02 Abstract

Mesoscale Effects of Climate and Land-cover Change on River-basin Hydrology on Amazonia

Thomas Dunne — University of California (US-PI)
Luiz Antonio Martinelli — CENA - Centro de Energia Nuclear na Agricultura (USP) (SA-PI)

Project Synopsis

Purpose of the Study

The goal of the proposed research is to answer the question: 

What is the

influence of land-cover change on the streamflow of Amazonian river basins with

areas of approximately 10,000 km2 ?

Since the effects of land–cover change at this ‘mesoscale’ interact

with the influence of the other major controlling variables, climate and basin

size, this question needs to be answered by accomplishing subsidiary tasks,

which are:

  1. to quantify the influence of both climate and land-cover change on river

    flow in Amazonian basins with areas of approximately 10,000 km2;

  2. to define the range of basin size over which land use, with its own

    characteristic nature and scale in Amazonia, affects the hydrology of

    rivers; and

  3. to understand processes governing those influences.

The results will be expressed in the form of a mathematical model, checked

against the response of river flow in the State of Rondônia, Brazil, and driven

by measurements of rainfall, energy sources, and terrain characteristics,

derived initially from ground-based measurements and eventually from satellites.

The model will be applicable to mesoscale basins in other physiographic regions

of the Amazon Basin, in which our research group has some field experience. The

purpose of the model is to assist in interpreting and generalizing from field

data and in understanding the processes by which climate and land-cover, along

with terrain characteristics, affect hydrologic response at the river-basin

scale. That understanding will be formalized through systematic computational

experiments, and checked against river discharges gauged by the Brazilian

hydrometric agency ANEEL. The empirical part of the study will define current

and recent effects of climate fluctuation and land use on river flow. The

intended level of understanding and modeling of river flows and basin hydrology

will facilitate sustainable management of water and land resources in heavily

exploited regions.



The rare opportunity exists in the state of Rondônia for a model-based

comparison of land-use effects on the hydrology of river basins varying from a

few hundred to 60,000 km2. Four subparallel, mesoscale basins: the

Jaci-paraná, Candeias, Jamari, and Ji-Paraná, run north to join the R. Madeira,

one of the major tributaries of the Amazon (Figure 1). These basins have been

gauged by the Brazilian hydrometric agency, ANEEL, for periods ranging up to

about 20 years. Their regional geology, soils, rainfall climates, and primary

vegetation covers (floresta aberta) are similar, but because the pattern

of land-use in the region is driven by the alignment of a major road, BR-364,

the approximate proportions of the basins that have been deforested at some time

between 1974 and 1996 are: essentially zero for the Jaci-paraná; <5% in the

Candeias; 51% for the Jamari, and 47% for the Ji-Paraná. Gauged tributary basins

within the Jamari and Ji-Paraná basins (Figure 1) represent a broader range of

deforestation, ranging up to 70%. The shaded zone in Figure 1 indicates the zone

of widespread deforestation that has occurred at some time before 1996.

It is not possible in such a large area and dynamic land-use situation to

perform a hydrologic experiment of the kind conducted in studies of land-use

effects in "paired watersheds". Important results on the microscale

response of forest and pasture in Rondônia will result from such a microscale

basin study and associated hydrometeorological observations to be conducted by a

Brazilian, British, and Dutch contingent in LBA. We have communicated closely

with this group throughout our planning, and hope to use results from their

towers and other sites to refine our energy and water balance computations.

However, that study does not speak to the mesoscale question concerning

river-basin response, outlined above.

In the absence of a controlled hydrologic experiment, we intend to define the

mesoscale land-use signal through a formal, model-based comparison of the

hydrologic response of basins with differing extents and histories of land-cover

change during the period of river gauging. We will base our modeling on field

measurements by ourselves and others to represent explicitly the effects of land

cover and other variables (mainly soil, topography, and channel network form) on

the physical processes of the hydrologic cycle (evaporation, infiltration,

soil-moisture storage, groundwater storage, subsurface flow, streamflow, and

floodplain inundation). At the mesoscale, these processes must be integrated to

a coarse-grained resolution, and this will be the task on which we will

concentrate. We will do this by characterizing relatively homogeneous Hydrologic

Response Units (HRUs) of approximately 10-100 km2 each using (1)

previously mapped and satellite-based information on terrain characteristics,

interpreted on the basis of our sample field transects of topography, depth to

bedrock and water tables, and subsurface hydrologic properties; (2) isotopic and

chemical tracing of runoff from HRUs; and (3) numerical modeling, and routing of

the predicted runoff to stream gauges currently maintained by ANEEL.

After comparing the results of the (un-calibrated) model with rainfall and

streamflow data from the gauges shown in Figure 1, we will conduct a set of

computational experiments to analyze systematically the influence of climate and

land use on the hydrologic functioning of mesoscale basins under a range of

conditions that is widespread in the heavily exploited portion of the Amazon Basin. Our emphasis will be on creating a general, process-based understanding

of the land-use signal on both the quickflow that generates floods and

transports bioactive materials into streams, and the delayed flow that sustains

water supplies and aquatic ecosystems, and reflects regional patterns of

evaporation, primary production, soil moisture, and groundwater recharge. Both

our systematic field surveys of variability and pattern in terrain

characteristics and our modeling will also facilitate the interpretation of

macroscale, coarser-grained approaches to observing and modeling the entire

Amazon Basin.

Figure 1: Map of

the four mesoscale rivers to be used in the study. Different colors indicate the

four major gauged river basins and the tributary basins nested within them.

Black symbols indicate the river gauges operated by ANEEL. The

approximate drainage areas gauged are: Jaci-paraná 13,500 km2 (light-blue

outline); Candeias 12,700 km2 (orange outline); and Jamari 14,800 km2

(dark-blue outline). Four gauged basins of approximately 1,000 km2

lie within and around the Jamari basin. In the Ji-Paraná (red outline), gauged

basins ranging in area from 3,000 km2 to 15,600 km2 have

been heavily affected by deforestation at some time during the past 25 years.

That river is also gauged at drainage areas of 31,000 km2 and 62,000

km2 downstream of the disturbed area, presenting an opportunity to

study the damping of the land-use signal by the hydrologic response of the

ambient undisturbed forested lowland and a major river corridor. The diagonal

axis of the mapped area is approximately 500 km long.

October 1, 1998


Close Window