The project is an attempt to do three different things simultaneously rather than consecutively:

• a) evaluate a broad spectrum of remote sensing and modelling techniques for the determination of yield indicators at local and regional scales

• b) identify constraints on the operational use of these indicators for timely crop yield assessment such as the availability of ancillary meteorological and agronomical data and the attitudes of users about these techniques

• c) initiate the integration of capabilities and of infrastrucutre towards reginal crop yield assessments for Latin America

WHY?

In Latin America agricultural production, both for export and for domestic consumption, relies heavily on precipitation. Even in those areas, such as western Argentina and central Chile, where large scale irrigation is made possible by melt-water accumulated during the winter in the upper reaches of the Andes, hydrological conditions are the main driver of agricultural production.

A combination of factors is driving the production system towards stronger competitiveness requiring more precise and timely management. Given the sheer size of thinly populated agricultural lands in Latin America, new ways of gathering timely information on agricultural production and hidrological conditions have to be devised.

Remote sensing and numerical simulation models still have a hi-tech flavour and an often-heard argument is that they are not relevant to the research and operational priorities of developing countries. This argument rests on several misconceptions, the worst being that new techniques cannot be imbedded in research on and management of land and water resources.

In many countries, such as Argentina, Brasil and Chile, reality could hardly be farther from this statement: not only there are small but active communities of specialists, but also the necessary infrastructure is being installed and run by research and operational organizations involved in agricultural research and related services, e.g. collection of meteorological data. This led to the idea of adapting to Latin Amercia the experience gained in Europe with crop yield assessment based on a combination of satellite data and numerical simulation models of crop growth.

HOW?

The project participants have pooled their expertise and theis on-going research projects. This offers the possibility of evaluating the same or similar techniques under a broad range of hycrological and farming conditions. The structure of the project reflects the three goals:

• a) research

• b) preparation for the operational collection and use of satellite data

• c) collection and organization of ancillary data

Four participants run their own ground stations for the real-time acquisition of satellite data; whereas the University of Chile has an established tradition in this field, INTA, CEPAGRI and EMBRAPA have established their ground stations during project execution and are facing their first operational problems such as defining data products and establishing an efficient processing chain.

Identification of users and of their requirements at this stage is timely and made easier by the experience of the other project participants. Sensors onboard satellites provide many useful measurements of land attributes, besides land cover and land use at different spatial scales. Numerical simulation models provide a practical way of integrating information, very diverse at times, and of comparing land use in different areas and conditions of develop practical strategies.

WHAT?

The project aims at developing pre-operational capabilities to receive, process and distribute useful data products to support research and applications dealing with early assessment of hydrological conditions and crop yield.

High resolution satellite data (Landsat TM and SPOT) are being used and new techniques have been developed to map land cover and land use: advanced classification techniques based on fuzzy logic and neural networks have been tested. Sub-pixel land cover has been mapped with AVHRR data at LAC resolution using spectral end-members.

Zones homogeneous with regard to the dynamics of green vegetation have been mapped using Fourier transform of extended time series (1981-1992) of AVHRR/NDVI measurements at GAC resolution. It has been demonstrated that SMMR 37ghz data are useful to assess the extent of flooded areas and storage capacity of natural reservoirs in large watersheds such as the Cuenca del Plata.

Satellite measurements of surface reflectance and temperature of spectral vegetation indexes have been used to map actual evaporation and the results successfully compared with a regional hydrological model. A joint field experiment has been carried out in the Altiplano to improve the accuracy of both forecasts and assessments of frost damage based on satellite thermal infrared observations. A crop yield simulation model has been applied to evaluate the potential of irrigated agriculture in Latin America, while other crop yield models are being evaluated against field experiments. A study on perceptions and preferences of users of agrometeorological data and indicators is being initiated in cooperation with the Central Statistical Office of Argentina. Field studies are being carried out in Argentina: irrigated lands in the west and rainfed intensive agriculture in the east; Brasil: Cerrado and São Paulo; Chile: central region; Bolivia: Altiplano; Uruguay: rangelands in the west and south.

[arriba]

The general objective of the project is to establish a cooperative network in South America for monitoring the impact of climatic and hydrological conditions on a agricultural production and increase the local capabilities and inter-relationships among institutions by means of exchanges of methodologies, results and training of their human resources in specific topics.

The general objectives are:

• Establish guidelines to improve adaptation of farming systems for varying climatic, hydrological and land use conditions

• Apply and assess accuracy of different methods to evaluate large areas of agricultural production

• Develop and test methods to derive hydrological indicators with remote sensing measurements

• Definition and development of pertinent data bases (agroclimatological, hydrological, statistical, satellite data)

• Assess perceptions and preferences about traditional techniques vs remote sensing based techniques of officers involved in management of agricultural production by means of marketing techniques

The project methodology to achieve the proposed goals is based on the use of recently elaborated technologies on crop modeling, including realistic water balance submodels, and remote sensing with the aid of earth observation satellites and meteorological satellites.

The project is divided in eight work packages each with specific objectives and tasks.

Work Package 0: Development of a regional cooperative network

Coordinators: Dr.M.Menenti, DLO-SC,NL and Dr.C.Rebella, INTA, Argentina

The main objective is the integration of the wide spectrum of disciplinary and national contributions into a consensus package of tools to support decision making in agricultural development and planning in the region.

A total of three annual workshops by research sector, a final general workshop and annual meeting of the project leaders, will be carried to identify the operational relevance of the project output. Each workshop will last for two weeks to allow cooperative work among participants for joint data analysis and evaluation of methods. Results of such workshops will be presented in a workshop proceeding series and reports for STD Programme Office.

The main topic and periodicity of these workshops are:

• Data production and management - annual

• Modelling - annual

• Satellite data analysis - annual

• Economic impact of advance information technology of agricultural development in South America - final year

Output: Periodic reports for STD programme office and workshop proceedings

Work Package 1: Generation of satellite data products

Coordinator: Dr.C.Rebella, INTA, Argentina

The objective is to generate high quality satellite data products, of agricultural and hydrological significance, using the raw satellite data from the ground stations of Chile, Brazil and Argentina. The major contribution will be to "jump-start" this activity by a many-fold increase in demand for such products and by providing additional resources for a limited period of time to the operators of these ground stations.

The general approach ti achieve the objective is through:

1. Time series analysis of existing low resollution satellite data sets (AVHRR/GAC, SMMR, SMM/I)

2. Acquisition of AVHRR/LAC and production of high resolution products (surface temperature, albedo, vegetation index)

3. Adaptation of the SPACE package to be provided by JRC Ispra (MARS Project)

Output: Variety of satellite data products

Work Package 2: Land use mapping

Coordinator: Dr.C.Rebella, INTA, Argentina

The objective is to evaluate several procedures, using high and low resollution satellites, for land useing mapping of large and small areas of agricultural production.

The methodological approach is based on:

1. Evaluation of spectral end-members and sub-pixel estimation using NOAA/AVHRR data

2. Technical and economical intercomparison of land use mapping techniques using Landsat /TM and Spot data

3. Land use mapping of two regions: Antiplano Region and Pampean Region

Output: Land use maps of different areas; guidelines for cost effective use of satellite images

Work Package 3: Evaluation of agricultural and hydrological indicators by means of time series analysis of meteorological satellite data sets

Coordinator: Dr.M.Menenti, DLO-SC, The Netherlands

The objective is to obtain agricultural and hydrological indicators by studying the evolution in time seriesprocesses, of meteorological satellite data sets, over long periods of time and extense regions.

The main activities are:

1. Fourier analysis of time series of surface temperature, albedo, vegetation index and polarization difference

2. Analysis of amplitude and phase images for eight test sites

3. Evaluation of AVHRR time series to monitor agricultural land in the

• Altiplano

• Rangelands in Mendoza

• Mata Atlantica in Brasil

• Agricultural land in the 5th Region of Chile

• Agricultural land in Paraguay and Uruguay

4. Use of AVHRR time series and Fourier transform to study the reaction of vegetation patterns to natural hydrological variability vs human impact

5. Development and test of advance classification procedures based on time series analysis of satellite data

6. Evaluation of time series analysis of satellite data and numerical models as tools to map bioclimatic units

7. Analysis of hydrological conditions in the Plata Basin using SMMR and SMM/I microwave observations

8. Mapping of freezing-prone areas with time series of AVHRR/TIR data

Output: Maps of a variety of climatic, hydrological and agricultural indicators based on time series of satellite observations

Work Package 4: Land surface energy balance and evaporation mapping

Coordinator: Dr.B.Seguin, INRA, France

The objective is to undertake a systematic evaluation of methods and to produce maps of relative evapotranspiration for the study of eight test regions selected by project participants.

The principal activities are:

1. Selection of ET-mapping algorithms; use of multispectral satellite observations

2. Application and evaluation of the algorithms in eight test sites

3. Validation with the hydrological model of Tunuyan irrigation district

Output: ET algorithms applicable to different physiographic regions; maps of actual evapotranspiration

Work Package 5: Evaluation of agrometeorological models

Coordinator: Dr.F.Santibañes, UCH, Chile

The objective is the adaptation of numerical simulation models of crop growth and farming practice to three test regions in Chile, Brasil and Argentina.

The methodology comprises:

1. Calibration and validation of agrometeorological models

2. Application to estimate agricultural production

3. Definitions study on the operational use of agrometeorological models

Output: Detailed design of agrometeorological information system; maps of potential agricultural production

Work Package 6: Perceptions and preferences of user satellite-based agrometeorological products

Coordinator: Dr.M.Menenti, SC-DLO, The Netherlands

The main objective is to identify and describe in detail two small sets of satellite products: a first set of agrometeorological products and a second one of primary satellite measurements. Product attributes are specified in the former case in terms of applications and in the latter case in technical terms.

The methodology will be based in surveys directed to two separate communities of agricultural management and of space science and applications. The principal activities are:

1. Preliminary study to define agrometeorological products for the general public (private and public sector) and satellite products for professional user (research and development sector)

2. Design and preparation of questionare and survey, execution of the survey

3. Data analysis, ranking of product attributes as regards perceptions and preferences

Output: Guidelines for product improvement and targeted marketing

Work Package 7: Collection of agrometeorological, hydrological and statistical data and set up of a data-base

Coordinator: Dr.H.S.Pinto, CEPAGRI-UNICAMP, Brasil

The objective is to set up and compatibilize a regional data-base of agrometeorological, hydrological and agriculture statistical data.

The activities involve:

1. Collection of information for sites:

• i) Pampean Region - Argentina

• ii) Mendoza Region - Argentina

• iii) South East States of Brasil

• iv) Altiplano Region - Bolivia

• v) 5th Region of Chile

• vi) Mata Atlantica Region - Brasil

• vii) States of Uruguay

• viii) States of Paraguay

2. Design of data base and functions

Output: Digital information system for agricultural planning

[arriba]

1. SC-DLO - Winand Staring Centre, Wageningen, The Netherlands

2. INRA/BIO - Institute Nationale de la Recherche Agricole, Station de Bioclimatologie, Avignon,Francia

3. TERMO/UVALE - Facultad de Fisica Universidad de Valencia, Dept.Termodinamica, Burjasot, Spain

4. EMBRAPA/CPAC - Empresa Brasileira de Pesquisa Agropecuária, Brasília, Brasil

5. UNICAMP/CEPAGRI - Universidade Estadual de Campinas, Centro de Ensino e Pesquisa em Agricultura; IAG/USP - Instituto Astronômico e Geofísico, Universidade de São Paulo

6. SENAMHI - Servicio Nacional de Meteorologia e Hidrologia, La Paz, Bolivia; ORSTOM - Office de Recherche Scientifique et Technique d'Autre Mer

7. FCA-UCHILE - Facultad Ciencias Agricolas, Universidad de Chile, Laboratorio de Agroclimatologia, Santiago, Chile

8. INCYTH - Instituto Nacional de Ciencia y Tecnica Hidrica; CRICYT - Centro Regional de Investigaciones Cientificas y Tecnologicas, Mendoza, Argentina

9. INTA - Instituto Nacional de Tecnologia Agropecuaria, Centro de Investigaciones en Recursos Naturales, Instituto de Clima y Agua, Buenos Aires, Argentina

10. DGRNR - Ministerio de Ganaderia Agricultura y Pesca, Direccion General de Recursos Naturales, Direccion de Suelos y Agua, Montevideo, Uruguay

11. DMH/DINAC - Direccion de Meteorologia e Hidrologia, Direccion de Aeronautica Civil, Ministerio de Defensa, Asunción, Paraguay