WCC-3 Working session WS - 10
Title: Climate, land degradation, agriculture and food security
Short Title: Climate and Food Security MANAGING CLIMATIC RISKS FOR ENHANCED FOOD SECURITY: KEY INFORMATION CAPABILITIES
‘WHITE PAPER’ on
12 34567R. Balaghi, M-C Badjeck,D. Bakari, E. De Pauw, A. De Wit, P. Defourny, S. Donato, R. 8191073Gommes, M. Jlibene, A. C. Ravelo, M.V.K. Sivakumar, N. Telahigue, B. Tychon
1 Institut National de la Recherche Agronomique, Division Scientifique, Département de
l'Environnement et des Ressources Naturelles, B.P. 415 - Rabat R.P., Morocco.
2 The WorldFish Center Jalan Batu Maung, Batu Maung, 11960 Bayan Lepas, Penang,
Malaysia & the Overseas Development Group, University of East Anglia, Norwich, NR4 7TJ,
UK
3 Université de Liège, Faculté des Sciences, Département des Sciences et Gestion de
l‘Environnement, B 6700 Arlon, Belgium.
4 International Center for Agricultural Research in the Dry Areas, GIS Unit, P.O. Box 5466
Aleppo, Syria.
5 Alterra – Wageningen UR, Centre for Geo-information, P.O. Box 47, 6700 Wageningen,
The Netherlands.
6 Université Catholique de Louvain, Département des Sciences du Milieu et de
l‘Aménagement du Territoire, Unité d‘Environnemétrie et de Géomatique, Place Croix du
Sud, 2 bte 16, 1348 Louvain-La-Neuve, Belgium.
7 International Fund for Agricultural Development, Global Environment and Climate Change
Unit, Via Paolo di Dono 44, 00142 Rome, Italy.
8 Food and Agricultural Organization, Viale Delle Terme Di Caracalla, 100153 Rome, Italy.
9 Centro de Relevamiento y Evaluación de Recursos Agrícolas y Naturales, Facultad de
Córdoba Agropecuarias, Universidad Nacional de Córdoba, C.C. 509, Córdoba (5000),
Argentina.
10 World Meteorological Organization, 7 bis Avenue de la Paix, P.O. Box 2300, 1211 Geneva
2; Switzerland.
1
Recommendations
1) Despite numerous scientific, technological and humanitarian efforts to address the issues of crop productivity, food insecurity in developing countries remains a critical concern. Since climate variability is a dominant factor influencing food production and food insecurity, concerted efforts should be made to factor climate information and climate risk management into the strategies to enhance food security.
2) Relevant and country adapted information products should be prepared to advise decision makers at the government and regional levels of the existence of powerful tools to manage climate resources and the associated risks, stressing their characteristics in terms of costs and benefits, as well as their potential to improve food security. Climate and agricultural data information should be processed in such a manner there are directly serviceable by the final users (farmers, decision makers, NGOs, education, etc) through dedicated models and other tools.
3) Farmers in developing countries should access to products that increase crop production and reduce climatic risks(advice, structural and non-structural mechanisms, such as insurance). These products could be derived from the location-specific processing of weather and agricultural data or come from global weather information networks. This implies that local data collection, their transmission, processing in simulation models and their dissemination to the farming community should all be improved in a coordinated manner as part of an integrated advisory and warning system.
4) The most relevant agronomic and economic time horizon for the strategic planning of farming is one to five years and since this time horizon is between seasonal forecasts and climate change scenarios, efforts must be made to enhance research efforts and financial inputs to address this issue.
5) ―Hotspots‖ for climate applications should be identified based on a global assessment of the vulnerability at upmost scale for food security where climate forecast skills are high and where capacity exists to use climate information to manage risks.
6) Tools and methods should be developed and disseminated to make available detailed agroclimatic reference material (climatic risk maps, crop distribution maps etc) at a scale that is useful for local planning (village to district). The word "detailed" refers not only to the geographic scale, but also to the thematic resolution, such as local crops and breeds and farming practices.
7) In order to realize the potential value of seasonal climate forecasts in agriculture, linkages between producers of climate information and applications and various end users should be enhanced through appropriate mechanisms such as capacity building for intermediaries and end users and strengthening institutional partnerships (meteorological, agriculture, remote sensing and statistic administrations in particular), especially in developing countries.
2
Summary Food security is expected to face increasing challenges from climatic risks that are more and more exacerbated by climate change, especially in the developing world. This document lists some of the main capabilities that have been recently developed, especially in the area of operational agro-climatology, for an efficient use of natural resources and a better management of climatic risks. Many countries, including the developing world, now benefit from well trained staff in the use of climate data, physical and biological information and knowledge to reduce negative climate impacts. A significant volume of data and knowledge about climate-agriculture relationships is now available and used by students, scientists, technicians, agronomists, decision-makers, and farmers alike, particularly in the areas of climate characterization, land suitability and agro-ecological zoning, seasonal climate forecasts, drought early warning systems and operational crop forecasting systems. Climate variability has been extensively modelled, capturing important features of the climate, through applied statistical procedures, agro-climatic indices derived from raw climatic data and from remote sensing. Predictions of climate at seasonal to interannual timescales, are helping
decision makers in the agricultural sector to deal more effectively with the effects of climate variability.
Land suitability and agro-climatic zoning have been used in many countries for
agricultural planning, thanks to the availability of new and comprehensive methodologies; developments in climate, soil and remote sensing data collection and analysis; and improved applications in Geographic Information Systems. Drought early warning systems are available worldwide at both national and international levels. These systems are helping decision-makers and farmers in taking appropriate decisions to adapt to short term climatic risks. Also, operational crop forecasting systems are now becoming available at the regional and national levels. In some developed countries, several efficient and well tested tools are now available for optimizing on-farm decisions based on the combination of crop simulation models and seasonal forecasts. However, in developing countries few tools have been developed to efficiently manage crops at the farm level to cope with climate variability and climate risks. Climate change impacts on agriculture and food security have been assessed in international studies using specific and efficient methodologies and tools. Adaptation to climate change and variability can also be facilitated through effective planning and implementation of strategies at the political level. The role of technological progress, risk transfer mechanisms and financial instruments and their easy accessibility to rural people are critical elements of climate risk management.
3
1. Introduction
The discussion of the links between climate and food security must take into account the four 1dimensions of food security, availability, access, stability and utilization. All of them are
somehow climate dependent. Availability of food refers to the actual production of food, which in turn depends on efficient use of resources such as crop varieties, land and water; availability of inputs and management skills; and competition from the use of the same resources from other sectors such as livestock, fisheries, etc. Access to food refers to people‘s
economic ability to access food as well as their ability to overcome barriers that stem from physical remoteness, social marginalization or discrimination. It also depends on people‘s
access to the resources that sustain agricultural production, particularly land and water, the agricultural technologies and financial services, and the markets for agricultural inputs and produce (IFAD, 2007). Stability refers to the continuity over time of availability and access of food supplies. Stability can be threatened by erratic climate, economic and political factors and several changes that gradually affect agricultural activities (e.g. land use, loss of labour, increasing prices, etc.). Utilization of food refers to people's ability to absorb nutrients. This is closely linked to health and nutrition factors.
This document lists some of the main capabilities that have been recently developed, especially in the area of agricultural climatology, a scientific field that combines the knowledge of agronomy and climatology, in order to understand the complex mechanisms by which climatic resources (mainly heat, solar energy, rainfall) are processed into crop production. Efficient management of climatic variability and associated risks requires that these complex mechanisms are well understood and modelled by the scientific community in order to develop decision support tools for decision makers, agronomists and, most importantly, for the farmers. Developments in communications and electronic media, in particular the ever-expanding cyberspace linkages through Internet and World Wide Web are changing the way farmers view information dissemination and exchange.
This document lists some of the capabilities available to practitioners and decision makers, starting with the dissemination of agro-climatic data analyses and advice. The next section covers the characterization of the climatic, environmental and agro-ecological resource base, which is a necessary step in order to quantify agro-climatic resources, plan for their optimal use and describe climatic risk patterns for crop insurance and long-term agricultural and food security planning. Developments in seasonal climate forecasting and their applications are described in the following section. This is followed by a section which deals with crop simulation models and satellite technology for crop monitoring and early warning systems. It covers two types of applications: the well established ground-based agro-meteorological techniques and remote sensing
2. The subsequent sections describe the tools available to assess
and forecast impacts of climate variability and change to improve tactical planning, and the technological progress, especially in information dissemination through Internet, and local knowledge as key elements to adapt to climate change. The final section focuses on the role of institutions and governance in planning for adaptation to climatic risks.
2. Making agro-climatic information available to users A significant volume of knowledge about climate and agriculture is currently available and to
convert this knowledge into action, it must be communicated to various types of users,
including scientists and technicians to users involved in operational aspects of agriculture, i.e.
1 Sometimes referred to as ―the four pillars of food security‖. 2 The term generally refers to the use of imaging sensor technologies including: instruments on-board aircraft and spacecraft as well as those used in spectroradiometers.
4
those dealing with production, storage of products, trading, etc. Communication about all
weather-dependent aspects of crop and animal production, food and non-food forest products,
as well as fisheries, can help improve food security or incomes through the exchange of
―messages‖ (data, information, knowledge), with feedback, between a ―producer‖ and a
―target‖ or ―audience‖. Types of audiences (clients) vary a lot and the messages must be
customized and refined by experience to achieve maximum impact. This also applies to the
communication media. Messages can vary from awareness creation and advocacy to on-farm
management advice, warnings, knowledge and information useful for planning at the level of
individuals, institutions and government. Efficient communication relies on reliable and up-
to-date data and information. Use of indigenous knowledge can lead to an easier adoption of
the message. Modern communications technology, including the Internet and wireless
telephones, offer potential to improve climate communication and data use, such as the
establishment of Farm Adaptive Dynamic Optimization (FADO) schemes. FADO is based on
the real-time collection of on-farm information such as weather and phenology and the off-
site processing of the information in order to derive farm management options that are fed
back to the village.
Many countries, including developing ones, now benefit from staff well trained in the use of
climate data, information and knowledge to reduce negative climate impacts on the above-
mentioned ―four pillars of food security‖, and to make better use of climate resources. Next to
universities, much training is dispensed by specialized schools operated by national
meteorological services. Of particular relevance are regional centres, some of them
established thirty years ago, which continue to train technicians, engineers and scientists. One
of them is the Regional Training Centre for Agrometeorology and Operational Hydrology
(AGRHYMET) in Niamey, Niger, which was established by the Permanent Inter-State
Committee for Drought Control in the Sahel (CILSS) following WMO Expert Missions in
1972 in response to the Sahelian droughts.
3. Characterization of the climatic, environmental and agro-ecological
resource base
Climate is now mostly regarded as a hazard, due to the political visibility of climate change
and media of atmospheric extreme events. However, there is a lot to be gained from looking
at climate not only as a natural hazard, but also as a ―resource‖. Resources must be known,
assessed in quantitative terms and properly managed if they are to be used sustainably, and
climate is no exception (Gommes and Fresco, 1998). Magalhaes (2000) argues that climate
should be treated as a component of the natural capital endowment of the region and as a
factor that may trigger crises that impact people, economic and social activities, and the
environment. It remains that climate is the first natural resource (Walker, 1815; Bernard, 1992) as it provides water, heat, and solar energy, without even mentioning many benefits
such as wind pollination and wind power. However, unlike soil and other natural resources,
most climate resources are variable over space and time, which is the source of the risk
component inherent in climate. For this reason, climate variability has been modelled thanks
to agro-climatic indices, statistical procedures and local knowledge in order to capture
average patterns. The strong impact of weather on crops in the world lead to the development
of locally adapted agro-climatic indices. One of these indices is the well-known Penman-FAO
index (Smith, 1991; Allen et al. 1998) which is strongly related to crop yields in many arid
and semi arid regions of the world. Recently, satellite imagery development allowed the
derivation of new agro-climatic indices from vegetation reflectance measures which are better
related to crops in many cases, particularly in arid and semi arid regions. The Normalized
5
3-sensor, is one
of these satellite indices. NDVI has been extensively used in vegetation monitoring, crop yield assessment and forecasting (Balaghi et al. 2008; Benedetti and Rossini, 1993; Girma et Difference Vegetation Index (NDVI), as registered since 1980 by the AVHRRal. 2005; Quarmby et al. 1993). Most of statistical procedures rely on relatively simple frequency analysis of time series at dekadal, monthly and annual levels as well as spatial interpolation. However, statistical analysis of climate remains an expert knowledge art which is very specific to local conditions.
As with climate, the characterization of all environmental categories that include climate is variable in space and especially in time as well. For instance, climate classification maps and agro-climatic suitability maps describe the ―usual‖ or ―average‖ conditions. It may even
happen that ―average‖ never occurs in practice, e.g. in climates characterized by the bimodal distribution of variables. For instance in some areas at the border between temperate weather systems and monsoon systems (e.g. part of Southern Africa), the average climatic conditions seldom occur. Similarly, the reliability of ―first rainy season‖ and ―second rainy season‖
(separated by a dry season) that characterize many climates is very variable. In practice, either the first or the second dominates, which makes agricultural planning very difficult and often results in not so intuitive cropping patterns developed over the centuries by farmers to minimize risk (e.g. planting at the end of the first rainy season; Gommes, 1985). A systematic effort in land-use planning is an appropriate way to assure sustainable agricultural development and efficient use of natural resources. Agro-ecological zoning (AEZ) is used to characterize geographic areas based on climate, soil, biological and yield information (Pascale and Ravelo, 1989; Luque, 2009; Ravelo and Abril, 2009). Agroecological zoning offers much scope for developing strategies for efficient natural resource management and in this context, recent advances in remote sensing and geographic information systems have made the task of integration and mapping of a wide range of databases much easier. There is, for example, a need to reduce the farmer‘s risk when
introducing a new crop. Both satellite and ground information are essential ingredients to develop advisory systems and planning strategies for new crop farming investments. The environments represented by agro-ecological zones are often associated with distinct farming systems and land-use and settlement patterns. Agro-ecological zones maps have been used in many countries for different agricultural planning applications, ranging from the physical location of research stations, the introduction of particular crops, cultivars and technologies to suit the conditions in different areas, the allocation of water resources to agriculture, fertilizer recommendations, policies and regulations for rural land use, inputs and technology subsidies, and others. These applications illustrate the attractiveness of the AEZ concept to planners and decision-makers of different stripes and colours: the bird-eye view of agricultural potential and constraints offered by integrating the key components of the agricultural environments is much easier to understand than a stack of single-theme maps. Whereas in the past the manual integration of spatial data from different disciplines, at different scales and accuracies, was a major bottleneck in developing AEZ maps, GIS technology makes this now perfectly practicable. The feasibility of rapidly defining agro-ecological zones by the combination of climatic, land use/land cover, terrain, soil and other data using Geographic Information Systems (GIS) procedures has been demonstrated in the last few years through a number of regional and country studies. The integrating principle of the AEZ concept and the ease of linking AEZ mapping units to single-theme GIS layers,
3 Advanced Very High Resolution Radiometer. The AVHRR sensor is a broad-band, 4- or 5-channel scanning
radiometer, sensing in the visible, near-infrared, and thermal infrared portions of the electromagnetic spectrum.
6
4 analysis of
well-defined agricultural environments in relation to food security. It has been well
established that access to and stability of the natural resource capital - particularly natural
vegetation, climate, soil, irrigation water and biodiversity - are major determinants of the
resilience of rural livelihood systems against climatic risk. Understanding the underlying including climate risk maps, make it perfectly suitable for undertaking a SWOTcauses of vulnerability resulting from changes in the stability of the natural resource base
requires an integrated approach, which considers both the differences in agro-ecological and
socioeconomic characteristics between different areas. GIS-based themes of "agro-eco-socio-
economic" zones make a lot of sense in order to assess structural vulnerabilities of rural
populations to climatic and other resource-related risks to their livelihoods. Although thus far
little progress has been made in developing integrated spatial frameworks combining both
biophysical and socioeconomic themes, the feasibility of this approach has been improved
over the last decade thanks to the vast amount of climatic, soil, terrain, land cover and remote
sensing datasets that has been made available to the public at large.
Needless to say, more detailed and accurate analyses also require more detailed and accurate
data on weather, soils, land cover, etc. Many of the improved tools can avail themselves of
better data, including remote sensing and new sensors such as those used to measure soil
moisture of leaf wetness, a crucial variable in the simulation of disease impacts. 4. Seasonal climate forecasts and their applications
Year-to-year variability of climate significantly affects the agricultural fortunes of most
farmers. For example, the all-Australian crop value fluctuates by as much as 6 billion dollars
from year to year, and these fluctuations are highly correlated with seasonal ocean
temperature changes (Nicholls, 1985). Farmers have to take a number of crucial land and
water management decisions during the growing season, based on climatic conditions, and
sometimes these decisions have to be taken several weeks in advance.
The past two decades have seen significant improvements in the forecasting of climate
variability, based on advances in our understanding of ocean-atmosphere interactions. Such
improvements permit the development of applications that predict climate at seasonal to
interannual timescales, helping decision makers in the agricultural sector to deal more
effectively with the effects of climate variability (Sivakumar, 2006).
Until 20 years ago, seasonal climate predictions were based exclusively on empirical-
statistical techniques that provided little understanding of the physical mechanisms
responsible for relationships between current conditions and the climate anomalies
(departures from normal) in subsequent seasons. Mathematical models analogous to those
used in numerical weather prediction, but including representation of atmosphere–ocean
interactions, are now being used to an increasing extent in conjunction with, or as an
alternative to, empirical methods (AMS Council, 2001).
A wide range of forecast methods, both empirical-statistical techniques and dynamical
methods, are employed in climate forecasting at regional and national levels (WMO, 2003).
Empirical-statistical methods in use at various Centres include analysis of general circulation
patterns; analogue methods; time series, correlation, discriminant and canonical correlation
analyses; multiple linear regression; optimal climate normals; and analysis of climatic
anomalies associated with ENSO events. Dynamical methods (used principally in major
global prediction Centres) are model-based, using atmospheric GCMs, coupled Atmosphere–
Ocean GCMs (CGCMs), and 2-tiered models. Hybrid models, such as a simple dynamical or
statistical model of the atmosphere coupled with an ocean dynamical model, are not being
4 SWOT Analysis is a strategic planning method used to evaluate the Strengths, Weaknesses, Opportunities,
and Threats involved in a project or in a business venture.
7
used operationally by any National Meteorological and Hydrological Service (NMHS) at the present.
A recent trend is to examine the potential use of Regional Climate Models (RCMs). These are complex atmospheric models that only handle a relatively small region (approximately the size of Europe) but with far more resolution than is possible using present global models, and that use boundary conditions supplied by a pre-run of a global model (Harrison, 2003). It is hoped that outputs from such models will provide greater temporal and spatial detail than is available from the global models. Relatively cheap workstations, and even Pentium 4-equipped PCs, are all that is required to run a RCM, and a number of experimental systems are running in various countries with and without other numerical capabilities using boundary conditions supplied by a global centre.
In several regions of the world, interpretation and delivery of the climate prediction information has been promoted more through the development of Regional Climate Outlook Forums, initiated by WMO, NMHSs, regional institutions, and other international
organizations. It is a forum that brings together the experts from a climatologically homogeneous region and provides consensus-based, climate prediction and information usually for the season having critical socio-economic significance. These forums bring together national, regional and international climate experts, on an operational basis, to produce regional climate outlooks based on consensus agreement between coupled ocean–
atmosphere model forecasts, physically based statistical models, results of diagnosis analysis and published research on climate variability over the region and expert interpretation of this information in the context of the current situation (Berri, 2000). By bringing together
countries having common climatological characteristics, the forums ensure consistency in the access to and interpretation of climate information. Through interaction with sectoral users, extension agencies and policy makers, RCOFs assess the likely implications of the outlooks on the most pertinent socio-economic sectors in the given region and explore the ways in which these outlooks could be made use of. Regional agriculture and food security outlooks are now regularly produced based on the climate outlooks after the RCOFs in some regions. The first International Workshop on Climate Prediction and Agriculture (CLIMAG), held at the World Meteorological Organization (WMO) in Geneva, Switzerland in September 1999 (Sivakumar, 2000) considered a number of important issues relating to climate prediction applications in agriculture: capabilities in long-term weather forecasting for agricultural production; downscaling; scaling-up crop models for climate prediction applications; use of weather generators in crop modeling; economic impacts of shifts in ENSO event frequency and strengths and economic value of climate forecasts for agricultural systems. As part of the broader CLIMAG program, the Asia-Pacific Network for Global Change Research (APN) and START supported a multidisciplinary research project to assess the potential for seasonal climate forecasts to reduce vulnerability to climate variability in South Asia. By using a systems analytical approach in Southern India and Northern Pakistan, the project demonstrated how cropping systems management can be altered by adapting to the underlying climatic variability.
The Agricultural Production Systems Research Unit (APSRU) in Queensland has developed a software tool, ‗Whopper Cropper‘, to help predict the production risk faced by growers (Cox
et al. 2004). This combines seasonal climate forecasting with cropping systems modeling to help producers choose the best management options (Hammer et al. 2001). Farmers can investigate the impact of changing sowing date, plant population, nitrogen fertilizer rate and other variables.
8
5. Crop simulation models and satellite technology for crop monitoring
and early warning systems
5.1 Agricultural meteorology and ground-based approaches
Drought is largely a social construct representing the risk of agricultural activity being
substantially disrupted by spatial and temporal variation in rainfall and temperature (Botterill,
2003). A critical component of planning for drought is the provision of timely and reliable
climate information, including seasonal forecasts, that aids decision makers at all levels in
making critical management decisions. This information, if properly applied, can reduce the
impacts of drought (Wilhite and Svoboda, 2000). Drought early warning systems help
decision-makers and farmers taking appropriate decisions to adapt and mitigate climatic risks
well in advance. Thanks to early information, decision-makers can warn farmers well in
advance of likely drops in yields due to unfavourable weather conditions. Such systems are
available at the international level (GIEWS56, FEWS, etc.) and at the regional level 7(AGRHYMET – EWS). Depending on data availability, each country can develop its own
system, using, for instance, the FAO approach (Gommes, 1997), which aims to optimize the
combination of several kinds of data: punctual (meteorological data) or continuous (satellite
data), historical or real-time data, in order to achieve reliable and accurate yield forecasts
(Tychon et al. 2002). Today, some operational forecasting systems are available worldwide at country and sub-8national levels. Amongst the most important systems we can mention GIEWS and MARS
(Monitoring Agriculture by Remote Sensing) managed by the European Commission. These
forecasting systems are based on agro-meteorological models with various levels of
complexity and empiricism. Currently, most operational climate impact models use
mathematical techniques that were developed in the Seventies and Eighties.
Agro-meteorological models are widely used in the world to understand the crop response to
weather and soils (Landau et al. 2000). These models often need many other accurate input
data on weather, soils and crops that are not easily available, particularly in developing
countries. They do also require the fitting of many parameters, a difficult task, even more so
in developing countries. They were elaborated when remote sensing (RS) technology was at
the very beginning of its development. For example, both GIEWS and MARS try to improve
their yield forecasts by using the low resolution imagery registered by synoptic earth
observation systems, such as NOAA-AVHRR (active since around 1980) and SPOT9-10VEGETATION (since 1998) (see for example the MARS Bulletin for Morocco
http://mars.jrc.it/mars/Bulletins-Publications). RS is particularly useful in semi-arid regions
where the state of the vegetation shows high year to year variations in relation with weather
variability – as is the case for the Mediterranean countries which face high climatic risks due to drought and climate change (Balaghi et al. 2008).
Several efficient and well tested tools are now available for optimizing on-farm decisions
based on the combination of crop simulation models and seasonal forecasts. The tools apply at
farm level and where seasonal forecasts based on El Niño/Southern Oscillation Index (ENSO
Index) have good predictive power, e.g. in Australia. Whopper Cropper (Nelson et al. 1999)
5 Global Information and Early Warning System of FAO. 6 Famine Early Warning System of USAID (www.fews.net/). 7 Early Warning System. 8 Global Information and Early Warning System co-ordinated by FAO (www.fao.org/GIEWS). 9 Satellite Earth Observation System (www.spotimage.fr/). 10 Multi-spectral scanning radiometer (on board SPOT 4 and 5 satellites) acquiring images in 4 channels with 1
km spatial resolution.
9
was developed because climate and market risk threaten the efficiency and sustainability of cropping systems in the grain/cotton belt of northern New South Wales and Queensland where the semi-arid climate is extremely variable. The mean sorghum yield associated with a positive SOI phase for September/October is 1000 kg/ha greater than that for a negative SOI. Whopper Cropper is designed to provide distributions of crop yields that enable the likely impact of management options to be rapidly evaluated. It was developed using an iterative development process that involved extension professionals and the target user group (see http://www.bom.gov.au/climate/cli2000/rNelson.html). 11, etc.) are used directly in statistical models to forecast crop yields at large scale. However, most models can use RS data as input in various 5.2 Remote sensing stages of the modelling process (parameters, input or driving variable) and it has been demonstrated that the performance of the models can be readily improved when RS data are Actually indices derived from RS (NDVI, EVIcombined with crop models (Boegh et al. 2004; Bouman, 1995; Doraiswamy and Cook, 2005; Maas, 1988).
However, operational application of RS in agro-meteorological modelling systems for crop yield prediction is very limited today. Various reasons have played a role with regard to the applicability of RS data in agro-meteorological crop models. Difficult access to RS data in near real time has, up to recently, been one of the reasons. Pre-processing complexity and analysis have also surely played their role. However, one of the main obstacles so far has been the disparity in scale between the process (crop growth on small fields) and the type of satellite observing system that can be used operationally and economically over large areas with high temporal frequency. This basically means that satellite sensors which fit the operational constraints (operational, economical, and available) do not observe with high enough spatial resolution (usually 1 x 1 km
2) individual crop fields in many parts of the
World. This means that crop specific biophysical parameters are difficult to extract from these types of satellite data, which makes it difficult to use them in a crop-specific agro-meteorological model. Also, maybe the last reason why RS was not included in agro-meteorological crop models comes from the fact that most developed countries are located in the northern hemisphere where persistent cloud cover is a constraint to the use of remote sensing images.
In recent years, the advancement of satellite sensor technology has gradually improved the spatial resolution of polar orbiting satellite sensors that can cover large areas with high temporal frequency (such as MODIS1213 and MERIS). These sensors can now observe the
Earth with a spatial resolution of 250 to 300 meters with high temporal frequency (daily). This spatial resolution is still too coarse to observe individual crop fields in many parts of Europe. However, it is likely that there will be at least some pixels where the fractional coverage of a single crop within the pixel is high. It is therefore necessary to obtain so-called "vegetation continuous fields" (Hansen et al. 2003) also called Area Fraction Images (AFI) that can be used to find those pixels and extract crop specific biophysical parameters from them.
11 Enhanced Vegetation Index. 12 Moderate Resolution Imaging Spectroradiometer (http://modis.gsfc.nasa.gov/). 13 MEdium Resolution Imaging Spectrometer. MERIS is a programmable, medium-spectral resolution, imaging spectrometer operating in the solar reflective spectral range.
10