Jose A. Marengo — CPTEC - Centro de Previsão do Tempo e Estudos Climáticos (INPE) (SA-PI)

William James Shuttleworth — University of Arizona (US-PI)

The goal of this project is to explore methods

for simultaneously assimilating remotely sensed data (on land cover, surface

temperature and surface radiation) to improve the diagnosis of the Amazon

surface energy budget in CPTEC models. The research explores aggregation

techniques that can be used to incorporate the effects of the subgrid-scale (~1

km) heterogeneity that is resolved in the remotely sensed data at the (~ 30 km)

grid scale used in the mesoscale models.

Aggregation theory is based on the requirements

that area-average scalar fluxes must be conserved, and that the model used to

describe area-average land-surface-atmosphere exchanges at the grid-scale should

have the same form as the model used to describe such exchanges at the pixel

scale. Applying these requirements simultaneously to the equations describing

surface momentum exchange and surface energy exchange at both the grid and

subgrid scale leads to identification of relations between the parameters that

control exchange between the surface and a "blending height" in the

lower atmosphere at the two scales. In general, the resulting relationships

express the parameters to be used at the larger grid-scale in terms of weighted

summations of those that apply at the smaller pixel-scale. In principle, it

ought also to be possible to assimilate relevant subgrid-scale remotely sensed

information via aggregation algorithms based on this approach to improve the

estimates of grid-average surface fluxes.

Preliminary work is being carried out over the

Tennessee Valley in the USA, where the RAMS model coupled with the SiB2 model

and nested inside the Eta analysis fields is being run over a 256 by 256 km area

with a 1 km resolution. The calculated surface fluxes and near-surface

atmospheric variables are regarded as the "truth". Low resolution (30

km) model runs are then made to simulate the GEWEX mesoscale model grids and (1)

explore the value of the "aggregate" and "tile" hypothesis

for combining several different remotely sensed data streams; (2) determine

whether surface temperature for the cloud-free portion can be used to deduce the

soil moisture for the whole grid square; (3) test whether mesoscale circulations

compromise the effectiveness of these hypotheses; (4) test whether topography

compromises the effectiveness of these hypotheses. Model studies will be made in

two regions in the LBA domain (Rondonia and Santarem) that have different

heterogeneous land covers, and where there will be LBA field data for

calibration/validation. To further this research, data from the CPTEC Eta model

is being routinely downloaded. At present approximately 9 months of data are

available.

A CPTEC student, Luis G. Gonçalves de Gonçalves,

has been identified to carry out doctoral research at the University of Arizona

in support of this project and we are exploring options to obtain the required

funding. Eleanor Burke will visit CPTEC for 1 month in early 2000.

homepage: http://www.hwr.arizona.edu/sur_facpro.html#shuttleworth