APPLICATION OF A SOIL WATER BALANCE MODEL TO THE MERCOSUR AREA. J. Tomasella, J.A. Marengo M. Doyle and G. Coronel MAR DEL PLATA OCTOBER 2002

Mutual interaction of soil atmospheric processes and soil moisture has been the focus of several studies since the 80’s. That interaction can take place at different time and spatial scales (e.g. Entekhabi et al. 1996): at local scale, on the partitioning of latent and sensible heat fluxes; on the growth of squall lines, fortogenesis and local circulation at synoptic scales on the anomaly reinforcement by surface hydrologic processes at a climate scale.

Besides atmospheric circulation, knowledge of soil moisture is crucial in several biological and physical processes, for instance: to understand runoff generation on hydrological models; for estimating crop productivity, crop damage and biomass production. Which in turns influence economic activites like hydropower, navigation, human consumption, agriculture, biological functioning of ecossytems, etc.

Although soil moisture has been identified as an important feedback mechanism, measuring and modelling of soil water storage at a regional scale proved to be complex and difficult on a routine basis because: soil moisture is difficult and expensive to measure, data is rarely available for few spots, using different sensors and methodologies ; the scale of variability of soil moisture processes is generally few metres, making difficult to represent processes on a regional scale; lack of an interdisciplinary approach to study and integrate existent knowledge.

Since 1998, CPTEC produce experimental soil water estimation for assessing critical areas of the Brazilian Northeast semi-arid. The purpose of this effort is to provide guidance for governmental mitigation actions. The model integrates meteorological and pedological information, incorporating climate and soil variability at a regional scale.

ETR/ETP S (mm) PM (1500 kPa)c=1-p (60 kPa)CC (20 kPa) S t+1 =S t +PRE-ETR 0 1 DéficitExcesso The model is based on a simple bucket that uses rainfall and potential evaporation as input variables. meteorological data (from the synoptic network) to calculate potential evaporation.

Potential transpiration is estimated using Penman-Monteith equation, according to FAO methodology (Smith, 1991): r c = 69 s m -1

NDVI Out2001-Jan2002 Fev2002-Maio2002 Jun2002-Jul2002

In mid 2001, an effort to apply and validate the model for Sourthern Brazil and Argentina begun under PROSUR. Because soil data was available for Buenos Aires province, the region was selected as a pilot area for model application. Meterological data comes form the National Weather Service of Argentina and several sources in Brazil.

Model parameters were derived from pedological data (soil surveys), using pedo-transfer functions

Pedo-transfer functions are statistical techniques for estimating the parameter of an analytical retention curve or the soil moisture at a specific potential from basic soil data, eg:

Knowing the soil water retention for each horizon of every soil profile, parameters such wilting point, field capacity, etc can be derived.  (kPa)  (m 3 m -3 ) WPcFC Those parameters can be integrated along the profile to the root depth.

Point information from soil profiles can be extended using interpolation techniques, incorporating (or not) grouping defined by soil maps.

In the pilot phase, soil water storage was estimated on a daily basis for a period

Rainfall anomalies La Plata River Basin (Source: CRU, Reference period ) EN-DJF 1982/83 LN-DJF 1988/89

Future plans: Apply the model in Paraguay and Uruguay (IAI support?). To monitor a larger area in Argentina. Validate the model using soil moisture na meteorological data in Brazil. To use CPTEC’s regional model forecasts to produce soil moisture estimations for application in agriculture To provide information for the low level jet experiment and initial conditions for atmospheric models. To contribute to the development of crop prediction models in Brazil.

FIN