Dr J.H.C. Gash, Institute of Hydrology, Wallingford, UK.

Dr A.J. Dolman, DLO Winand Staring Centre, Wageningen, Netherlands.

Dr J. Noilhan, Meteo France, CNRM, Toulouse, France.


Version 3.0, 27 June 1997

This document was approved at the European Workshop for the Large Scale Biosphere Atmosphere Experiment in Amazonia held in Wageningen, Netherlands 23-25 June 1997.






Participating institutions:

Institute of Hydrology

DLO Winand Staring Centre

Centre National de la Recherches MJ etJ eorologiques

Wageningen Agricultural University

Laboratoire de MJ etJ eorologie Dynamique

Institut fh ur Meteorologie und Klimaforschung

Vrije Universiteit Amsterdam







The EU contribution to the Physical Climate component of LBA has the following priority elements

The central theme of the contribution is understanding how the transport of energy and water through the atmospheric parts of their cycles within Amazonia will change in response to changes in land use and global climate forcing. Emphasis is on how the influence of the land surface on rainfall, changes with land use. This programme primarily addresses primarily issues from the local to the mesoscale.

Specifically we will:



The physical climate component of LBA will study the transport of energy and water through the atmospheric part of the energy and water cycles, and investigate how the interactions between the vegetation and the atmosphere influence these cycles. The tropical land and atmosphere form a highly coupled system. The surface fluxes not only control the inputs of water and energy to the atmosphere, but depend themselves on the dynamical and thermodynamical properties of the planetary boundary layer through a chain of processes involving cloudiness, soil water content, evaporation, sub-surface hydrology and vegetative cover. The Physical Climate component of LBA will observe and model these coupled land-surface / atmosphere processes to provide answers to the following questions:

1. What are the surface and meteorological controls on the fluxes of energy and water , and how do they vary both in space, over Amazonia, and in time, between seasons and from year to year, to affect the regional budgets of energy and water.

2. What are the mesoscale mechanisms by which differences in surface characteristics translate into large-scale weather and climate anomalies?

3. What is the role of dry and moist convection in transferring energy and how will it change with different land use patterns?

4. How is the rainfall of Amazonia controlled by the large-scale land-surface-atmosphere interaction? Which areas within Amazonia have the most influence on rainfall and how does this vary in time?

5. What is the relative importance of Amazonia in generating its own climate compared to the role of external planetary-scale forcing, and conversely what is the influence of Amazonia on global climate?

6. How will the climate of Amazonia change in response to changes in land use and global climate forcing?

The programme designed to address these issues can be broadly divided into three highly inter linked components. Long term monitoring, atmospheric mesoscale campaigns and modelling.

The proposed European contribution to the physical climate component of LBA will emphasize the issues 1-4, and touch upon 5-6and 6. This contribution is outlined in the current document which was finalised during the European Workshop for the Large Scale Biosphere Atmosphere Experiment in Amazonia held in Wageningen, Netherlands 23-25 June 1997.




2.1 Background

A long term field measurement programme consisting of a number of sites, well separated along the eco-climatological gradients within Amazonia, will be implemented. The European Physical Climate contribution to LBA aims to establish and operate three four of the LBA long term monitoring sites: forest and pasture sites in Rond^ onia, and a forest site at Sn ao Gabriel in north west Amazonia. It will collaborate with other groups at sites which are funded under Brazilian, US and other EU components (see programmes for Carbon and Hydrology components for EU funded measurements in flooded forest and in central Amazonia) with respect to the physical climate aspects of the measurements. Measurements will be made of standard weather ? and soil moisture variables and the surface exchanges of CO2 , energy, and water vapour. Measurements of plant physiological and soil properties which control the fluxes will be made during campaigns based at these tower sites. The data from these sites will be used to determine the main surface and meteorological controls on the magnitude and cycle of the fluxes of energy, water vapour, carbon dioxide and momentum and will provide calibration and validation data for one-dimensional flux models. Model sensitivity analysis will investigate how the sustainability of the ecosystems will be affected in a changed climate. The sites will be run in close collaboration with Brazilian local counterparts.






2.2 European contribution to long term monitoring


Long term site

Start date

Source of set-up costs

Source of running costs

Rondo^ nia



Tranche 1 PP-G7

EUSTACH-LBA and this proposal

Rondo^ nia



Tranche 2 PP-G7

EUSTACH-LBA and this proposal

Sn n ao Gabriel forest



EUSTACH-LBA and this proposal


Table 1. Long term monitoring sites to be operated by the European contribution to the physical climate component of LBA.



3.1 Background

Several studies have shown that mesoscale land surface variability can influence both the amount of precipitation and its spatial distribution. These changes can be caused by changes in horizontal convergence and associated vertical velocities, and rapid evaporation from the vegetation canopy and bare soil. This influence of landscape variability on rainfall at the mesoscale is an important aspect of predicting future Amazonian climate, as deforestation generally takes place in a piecemeal fashion with the resultant landscape being a dynamic mix of original forest, clearing and regrowth. The proportion of the original forest which is removed, or the spatial pattern of the deforestation thus may have an important bearing on the future climate. A programme of measurements and modelling within the EU component of LBA PC has been designed to study the effect of mesoscale surface variability on rainfall and mesoscale climate. This programme will specifically address Pphysical Cclimate iIssues 2 and 3.

The EU contributionís focuses is primarily on the understanding of the effects of land use change on the climate from the local to the mesoscale. The measurements to be made during the dry season Atmospheric Mesoscale Campaign (AMC) largely reflect the deficiencies of current mesoscale models. These appear to be the neglect of aerosols in the radiation and thermal budget equations, and the possibility of enhanced turbulence over the pasture areas. The AMC will provide the data needed to initialize, calibrate and validate the models. In addition the wet season AMC has been designed to study and the interaction of the land surface with convective processes' complementing the TRMM satellite mission and the study of the representation of large-scale, organised convection in GCMs.




3.2 Wet season

The first AMC will be in the wet season of 1999 and will tie in with the TRMM satellite mission described below.

Moist convection is the main route by which water, energy and trace constituents move from the surface layer of the atmosphere into the troposphere and effort will be directed at understanding this process more fully. Much of the impact of changes in land use on rainfall and climate is expected to occur through changes in the intensity and frequency of convective storms. There is at present no data set of tropical moist convection over land. Thus the measurements, particularly from the wet season AMC, will be of great value for validating the cloud and convection schemes of all General Circulation Models.

Data from the dense observation network put in place for the wet season AMC will be studied to characterise the large scale convective systems which propagate towards the equator across the Rondo^ nia region. These large scale organised systems are of a large enough scale to be resolved by GCMs and it is important that their properties and behaviour are fully understood if they are to be represented correctly. The convective systems will be characterised (size, height, life cycle, optical properties, trajectory etc) by using measurements from geosynchronous satellites. When a system enters the network of radio-sondes or is seen by TRMM or SSM/I, the thermodynamic and dynamic perturbations will be investigated in detail. The large scale dynamics in which the systems are embedded will be identified using the ECMWF analysis or reanalysis. Statistics on the results will be performed depending on the characteristics of the systems and on the large scale environment. These comparisons will enable energy exchanges of the convective systems with their environment and the surface to be described and quantified. This will enable the ability of forecast models (ECMWF) and climate models (LMD) to simulate the energetics of these systems to be evaluated. The synergy of the various observing systems will allow the evolution of the systems through their entire life cycle to be described. This will be an important contribution towards understanding the mechanisms governing the frequency and intensity of events. However, the current planing of in-situ observations would have to be augmented by more frequent radiosondes to the west of the AMC domain to obtain a detailed description of large scale convection at a given point in time and space.

Measurements of the fluxes of momentum and heat in the atmospheric boundary layer will be made during the wet AMC. These measurements will be made to determine whether the heterogeneous forest-pasture surface, leads to significantly different momentum and energy exchanges than those which occur over forest or pasture alone and if current, theoretically based models of the interaction of the atmosphere with heterogeneous terrain perform adequately. The flux measurements will be made by 3 turbulence probes, attached to the tetherline of a large helium filled (~100 m3) tethered balloon The probes will be positioned up to 500 m above the surface and will be situated at heights chosen on the basis of preliminary modelling studies. Inclination and heading data will also be measured at high frequency (~20 Hz) and transmitted to allow corrections for probe motions to be applied to the data.

During the AMC the basic surface network of flux measurements over areas of forest, regrowth and pasture in Rondo^ nia will be augmented by extra sites at burnt pasture and possible an extra site at a forest. This is necessary to document the time and space variability of surface fluxes of available energy, sensible and latent heat, carbon dioxide, and momentum, over the various ecosystems and to detect mesoscale variability.

An array of radiosonde stations, jointly operated between Brazilians and EU groups will be operated at 4 sites, including tethered balloons to provide budget estimates of temperature and humidity. This set will allow estimates of moisture convergence to be made for the experimental area.

The following table summarizes the instrumentation proposed for the wet season AMC



RASS/Sodar system


Sodar system


5 radiosonde stations



Tethered balloons



Tethered flux balloon


Network of rainfall gauges


Table 2

A 2-D boundary layer model will be used to develop and test aggregation rules for the mixed forest - pasture areas in conjunction with the data from the tethered balloon flux profile. The measurements will be used to test for the existence of a blending height and to determine how this height is related to the length scale of the heterogeneity of the surface. If the blending height is shown to exist, then it and the properties measured at the surface over the individual components of the land surface will be used in existing aggregation parameterisations to make predictions for area average aerodynamic roughness and other parameters which can be tested directly against values derived from the tethered balloon data. New formulations will be derived if existing ones are shown to be unsatisfactory.

Mesoscale modelling work will be a continuation and development of work carried out under the CABARE project. The objectives are to investigate the effect of deforestation on the wet season rainfall patterns in Rondo^ nia. More specifically, to (font) to use the mesoscale modelling system RAMS to investigate the effect of forest conversion on boundary layer growth and to use RAMS to investigate the response of convection to particular patterns of deforestation in Rondo^ nia.

(US spelling here! modeling) RAMS (the Regional Atmospheric Modelling System) has been developed at Colorado State University and the ASTER Division of the Mission Research Corporation. It has had widespread use in simulating and forecasting meteorological phenomena. The two major components in RAMS are the data analysis and assimilation package and the atmospheric model. Two-way interactive grid nesting allows simultaneous modelling of large scale and small scale phenomena. The required steps in this modelling programme are, 1) pProduction of land use sets , 2) calibration of surface schemes, 3) assessing performance of precipitation scheme, 4) runs for selected days in wet seasons , 5) sensitivity studies for land use/degradation effects on rainfall and ultimately 6) predict longer term effects

The land use has been classified in an earlier project (Alvala et al., 1996) and calibrations for the land surface scheme of RAMS have been obtained for conditions during the dry season. tThe performance of the land surface scheme, particularly the interception component will need to be assessed. Satellite imagery will be used to assess cloudiness of the region. If possible estimates of rainfall by satellite will be used to provide a indication of regional rainfall patterns.


3.3 Dry season

The second AMC will be in the dry season of 2000. The dry season AMC will need a similar setup (except for the measurements required specifically for rainfall detection and TRMM). In the dry season campaign aerosol detection from aircraft will be critical. Validation Adjustment is being made with the aAtmospheric cChemistry component who will make these measurements for this purpose.

Surface based remote sensing systems will be employed for documenting the structure of the atmospheric boundary layer (heights of convective layer, mixed layer, entrainment zone, residual layer if any), and for observing the flow in the boundary layer with particular emphasis on investigating any local circulations which may be set up by the mesoscale variability of the surface. 2Two lidars will be deployed, along with the a RASS system (FUA) and one or more sodar systems.

Measurements of the fluxes of momentum and heat in the atmospheric boundary layer will also be made during the dry AMC.

4-5 radiosounding sites will be operated in close collaboration with the Brazilians scientists. During the AMC sondes will be launched at kleast 6 times per day from these sites to provide the atmospheric profiles required for the modelling studies. The data will be fed into the meteorological observing network to allow the data to be used in weather forecasts.

Dry model runs will be carried out using PERIDOT and Meso-NH. These models will be used to gain understanding of the interaction of forest and pasture in the cleared areas. The surface scheme in the models will be extended to include carbon dioxide. 3D simulations with Meso-NH will be carried out to document the possible effect of local breezes. Also, a new version of the Isba model (Isba-Ags) has been implemented. Isba-Ags is able to compute photosynthesis and leaf area index allowing comparisons of simulations with in situ measurements of carbon dioxide fluxes.

RAMS will continue to be used to understand the generation of increased turbulence by the stripes of forest in 2 and 3-D modes at high resolution, focusing on scales from 1 to 200 km. Collaboration will be sought wityh US groups to implement an aerosol parameterization in RAMS. Emphazsis will be on boundary layer development and effects of aerosols, and the generation of mechanical and thermal turbulence due to deforestation patterns.





The European contribution to GCM modelling will concentrate on two main themes. The first of these is uses the data gathered during LBA to improve climate simulations over Amazonia. Particular emphasis will be placed on the hydrological and energy balances at the surface, including their diurnal variations, and their sensitivity to parameterizations of convection.

The second theme is to study the sensitivity of the Amazonian climate to land surface modifications. The use of a stretched grid GCM (ARPEGE-Climat ) will allow simulations with a high resolution (50-100 km) in the vicinity of Rondo^ nia to be carried out thereby taking into account the geographical variations of land surface characteristics at this regional scale, as well has the orography in the neighbouring regions. This variable resolution technique, once it has been carefully validated with observational data, would enable the model to be used to perform sensitivity experiments to partial deforestation. Most of the deforestation simulations which have been performed with GCMs to date have used a coarse horizontal grid of more than 300 km size and extreme and unrealistic scenarios of total deforestation. The use of a finer grid in the region of interest would allow more realistic future deforestation scenarios to be tested. A comparison of the model response to the vegetation changes which have been observed in the last 20- 30 years with observed trends over this period could also be carried out.






All of the activities in the European contribution to the physical climate program of LBA will be carried out as a collaboration between European and Brazilian scientists (Table 3). The training aspect of this contribution to LBA will be based largely on the successful experience of ABRACOS in which the training programme involved a mixture of working closely with experienced scientists in the field and joint data analysis with scientists in the UK. Brazilian students, graduate students and experienced scientists will work together with European *researchers during planning, data collection and data analysis. There will be many *opportunities for Brazilian scientists to work at the appropriate European Institutes and 4-5 *several students are expected to register for PhDs related to this work.

* - should we expand this, or at least give number of PhD opportunities?



The timing of the activities proposed under the European contribution to the Pphysical Climate component of LBA is shown in the bar chart below. Data collection from the long term monitoring sites will continue for 5 years after the installation of the sites in 1997 or 1998. Data quality control and archiving will be carried out throughout the data collection period. Analysis and modelling of the data is likely to begin in 1999 and continue until 2002, with the largest scale and most data intensive modelling being concentrated in the last 2-3 years of this period. Some preparatory wet and dry season modelling modelling will take place earlier. The AMC will be held in early 1999 (wet season) and mid 2000 (dry season).




8.1 Carbon storage and exchange component of LBA

The four long term measurement sites in this programme will also make the measurements of carbon fluxes and storage required by the carbon storage and exchange component of LBA . Close collaboration will be established with respect to the PC components of these long term measurements.

8.2 TRMM satellite mission

The TRMM satellite, which is scheduled to be launched in November 1997 and will have a lifetime of three years. As part of the evaluation procedure there will be ocean and land field campaigns, which will be used to validate the models used for converting TRMM observations into rainfall. The wet season AMC in Rondo^ nia will be one of these campaigns. The extra instrumentation brought in by the TRMM team would include:

NASA ER-2 aircraft, equipped with Doppler radar, multi-frequency passive microwave sensors, electric field sensors, visible/infrared sensors (and more) for 12 -15 flights. The TRMM mission requires overflying major convective clouds and organised mesoscale precipitation systems.

Two Doppler radars. US scientists plan to propose bringing NCAR's S-POL radar, a state-of-the-art 10 cm radar with polarization diversity. The radar has a quantitative range of about 150 km. A second Doppler radar would be sited 30-50 km from the S-POL. This dual Doppler capability would be used to obtain 3-dimensional wind fields on the convective and mesoscale.

8.3 NASA Hydrometeorological module

The NASA hydrometeorology effort will concentrate on the use of remote sensing (aircraft and spaced-based) to study meso to basin scale issues. Close ties, both scientifically and functionally, with the TRMM validation efforts.

8.4 Modelling links

In addition to the European mesoscale modelling described above, other modelling work is proposed in a number of international centres. Although some overlap and reduplication, especially at the critical meso scale is likely to be useful, good coordination between the various modelling groups is required to achieve the general LBA modelling aims. CSU is interested in using RAMS to work with coupled atmosphere/hydrological models. Rutgers University is interested in nesting a mesoscale model (RAMS) in global models and working towards improving the understanding of the linkages of mesoscale weather systems and continental climate. In the case of RAMS joint parameterization programs may be established. USP and CPTECís interests are in the interaction of mesoscale with the large scale, the effect of the Andes and topography on the regional weather. Within the TRMM science team modelling plans exist which will use cloud ensemble models, radiative transfer models and mesoscale models (RAMS, MM5).

8.5 Aircraft observations

In 1995, CNRM, the Federal University of Cear< (UFC) and the University of Manchester (UMIST) started an experimental project for the study of precipitation formation in North-East Brazil, as part of the Int. Coop. Programme of the EC. The first step consists in the implementation of instrumental and modeling tools at the UFC. A field experiment is planned for the end of 1997 in North- East Brazil. The measurements will be performed with the FUNCEME Bandeirante aircraft equipped by the UFC/LFNM. Before the experiment, the new aircraft data processing software developed at CNRM by GMEI/MMA will be implemented at the UFC. The Meteorological Research Group at INPE and CPTEC have been recently involved in this co-operation project. There is now a possibility of modifying the structures of the INPE Bandeirante aircraft for implementation of instruments for in situ and remote sensing. This aircraft has been already used for air sampling and photographic measurements. Contrarily to the FUNCEME Bandeirante, it will be fully devoted to atmospheric research. The plans include complete thermodynamics (temperature and humidity high frequency rate measurements), dynamics (three components of wind and turbulence), three pods for particle spectrometers, accurate positioning of the aircraft, air and aerosol sampling and remote sensing windows for radiative measurements. The contributions of the INPE Bandeirante to field projects will be extremely valuable, especially for the LBA field campaigns.




9.1 EUSTACH-LBA (European Studies on Trace Gases and Atmospheric Chemistry as a contribution to LBA)

A research project in response to the EC-RTD 4th Framework Programme, Environment and Climate 1994-1998 this project aims to answer the following. What is the contribution of Amazonia to the global atmospheric carbon balance? What is the contribution of Amazonia to the global atmospheric budgets of radiatively active trace gases, and aerosol particles? and other trace gases which are chemically reactive in the global tropospheric ozone cycle? In relation to the EU Pphysical Cclimate contribution EUSTACH-LBA will contribute to the infrastructure and instrumentation at the Sn ao Gabriel Forest site and provide funding for site maintenance visits to Sn ao Gabriel, Rondo^ nia forest, Rondo^ nia Pasture and vVarzea flooded forest sites for 3 years, as well as staff time for data archiving, quality control and initial modelling.


  1. Tranche 1 PP G7 funding

WillTranche 1 PP G7 will fund the establishment of the Rondo^ nia forest site.


9.3 Tranche 2 PP G7 funding

Will Tranche 2 PP G7 will fund vehicles, equipment and running costs of the San o Gabriel site, equipment and infrastructure for the Rondo^ nia pasture site., dDevelopment of boundary layer balloon borne flux measuring instrumentation. and a Ppre-LBA wet season radiosounding campaign (RBLE 4).


See table 2.