Early Warning Systems

ITALIAN COOPERATION

Early Warning Systems
and Desertification

October 1999

Table of Contents

EXECUTIVE SUMMARY...................................................................................II

1 BACKGROUND............................................................................................. 1

2 INTRODUCTION .......................................................................................... 2

3 THE PRESENT SCENARIO ........................................................................5

4 EARLY WARNING SYSTEMS ...................................................................8
4.1. OBJECTIVES, PRODUCTS AND USERS ......................................................... 8
4.1.1. Objectives .........................................................................................8
4.1.2 Products..........................................................................................10
4.1.3 Users...............................................................................................11
4.2 METHODOLOGY....................................................................................... 13
4.2.1 Type of system.................................................................................13
4.2.2 Scale of application ........................................................................15
4.2.3 Indicators and thresholds ...............................................................16
4.2.4 Geographical distribution ..............................................................18
4.3 MECHANISM OF ACCESS TO INFORMATION .............................................. 22
5 INFORMATION DISSEMINATION SYSTEMS .....................................22

6 DATA DISSEMINATION SYSTEMS ....................................................... 24

7 EARLY WARNING SYSTEM AND DESERTIFICATION.................... 28

8 CONCLUSIONS........................................................................................... 29

I

EXECUTIVE SUMMARY

1 Background
In order to contribute to the preparation of the third session of the Committee on
Science and Technology and in accordance with decision 12/COP.2, the Italian Co-
operation and the Permanent Secretariat to the UNCCD have convened a four-day
workshop at the Agrhymet Regional Center in Niamey, Niger (from 25 to 28 October
1999).

The specific objectives of this initiative are:
1. to provide the Committee on Science and Technology with an evaluation of the
prospects of integrating early warning systems with environmental information,
focusing, in particular, on desertification.
2. to create an “enabling environment”for the development of early warning
systems on desertification, by:
- supporting the establishment of operational exchanges between existing
projects in the fields of early warning and environmental monitoring over
Africa;
- promoting the development and testing of practical examples for the
integration of early warning approaches with those of desertification.

On the occasion of this workshop, the CeSIA - Accademia dei Georgofili, on the
basis of the arrangements between the CCD Secretariat and the Italian Co-
operation, prepared a global report on Early Warning Systems and Desertification.
The executive summary of this report is annexed.

2 Introduction
Late in the ‘70s, as a consequence of the dramatic drought occurred in the West
and East Africa, the famine struk millions of people. The affected areas’
administrations and the international community were faced with the need to
provide the appropriate tools to facilitate the mobilisation of measures aimed at
mitigating the impact of recurrent droughts.

The EWSs that were conceived and implemented in that period can be considered
the ancestors of the systems in use today. In fact, from a methodological point of
view, they aimed at forecasting the occurrence of a risk situation on two scales: i)
the geographic area and ii) the population involved.

Since then, as a consequence of both the great results expected from the EWSs -
due to their theoretical potentiality as well as to the financial investments involved -
II

and the unimpressive results obtained, a passionate debate was generated at the
international level to find the most appropriate solution.

It was only in the mid ‘90s, when the technological evolution in telecommunications
(the Internet) and information technology (hardware and software) took place in
parallel, that the conceptual revision of the structural constraints of the technological
type enabled the parallel experimentation of new operational approaches.

Based on the above elements, the EWSs are still evolving, thanks to the
development of the vulnerability mapping systems, towards the integrated
management of the structural vulnerability analysis. The last is meant as the
capability of a population, a village, a social group to face a negative event, with
the risk forecast, that is the possibility that a negative event occur at a given time.

Since now, most of the operational EWSs are not addressing the environmental
aspects, desertification in particular, both in terms of indicators and of factors
affecting food security. At the same time, the systems dealing with the monitoring of
natural resources have given priority to the environmental aspects, leaving aside
the human being, as the affected and affecting element of the status of
desertification.

At present, the future scenario seems particularly favourable to a further evolution
of EWSs, as it is characterised by:
- an increasingly accessible and timely information;
- the development of data integration techniques aimed at producing an
immediate, useful and diversified information, according to the needs of
the different end users.

In fact, a complex and global system is being developed, formed by ‘ entities’
producing and distributing processed data, those which are more immediately
involved in early warning and those which create a favourable environment for the
circulation of information.

3 EWSs, Data Dissemination Systems and Information Dissemination
Systems: the present scenario
An EWS is based on three main components, namely:
the collection of data,
the processing of data and the production of information, and
the dissemination of information.

III

On the basis of this classification, a comparative analysis of the systems operating
in the Internet has been undertaken in the latest months. As a result of this, not so
many Early Warning Systems can be considered as such in the strict sense of this
expression. While the Information Dissemination Systems are still less numerous,
there is a definitely larger number of Data Dissemination Systems.

The systems inventoried are as follows:

A. Early Warning Systems:
1. Agrhymet Alerte Précoce et Prévision des Production Agricoles (AP3A)
project
2. USAID's Famine Early Warning System (FEWS)
3. SADC Food Security Programme (/REWU)
4. FAO Global Information and Early Warning System (GIEWS) on Food and
Agriculture
5. FAO Food Insecurity and Vulnerability Information and Mapping Systems
(FIVIMS)
6. WFP Vulnerability Analysis and Mapping (VAM)

B. Information Dissemination Systems on environment or desertification:
1. WB PRGIE
2. WB Environment Information Systems (EIS) in Sub-Saharan Africa
3. OSS System for the circulation of Information on Desertification (SID) /
Environmental Information and Monitoring System on the Internet (SISEI)
4. Scot Conseil and Medias-France Desertification Data and Information
System (D-DIS)
5. CEO - Desertification Information Network

C. Data Dissemination Systems:
1. European Spatial Agency (ESA) IONIA
2. Environmental Systems Research Institute (ESRI) Digital Chart of the World
(DCW)
3. EUMETSAT
4. FAO-AFRICOVER project
5. FAOSTAT
6. NOAA Satellite Active Archive (SAA)
7. NOAA/NASA Pathfinder
8. PENN STATE UNIVERSITY, Digital Chart of the World Data Server
9. UN Africa Nutrition Database Initiative (ANDI)
10. UNEP Global Resource Information Database (GRID)
11. USAID CARPE
12. USGS Africa Data Dissemination Center (ADDS)
13. USGS Earth Resources Observation Systems (EROS) Data Center (EDC)
14. USGS LANDDAAC
15. USGS Global Land Information System (GLIS)

IV

16. USGS Global Land Cover Characteristics (GLCC)
17. Centre for Environment and Development for the Arab Region and Europe
(CEDARE) GIS Database
18. WMO Global Climate Observing System (GCOS)
19. European Centre for Medium-Range Weather Forecasts (ECMWF)
20. World Conservation Monitoring Center (WCMC)- Forest Conservation

4 Basic caracteristics of EWSs
The end users of the early warning system should be the monitored populations
(target groups). However, in general, information is not directly reaching these
users and it is filtered through the national/local institutions providing them with the
most objective basis for the identification of the necessary actions to be taken.

Therefore, the final objective of any EWS is to provide the decision-makers with the
necessary and timely information on the present food situation in the relevant areas,
and the forecast for the end-of-season. As far as objectives and products are
concerned, every single system is characterised by differences depending on the
geographic area of application. These differences are often due to the availability of
financial resources, the availability of data in the national setting, and to the specific
agency requirements for the structure and the contents of outputs.

The most modern EWSs are based on a very extensive and multidisciplinary
analysis. The socio-economic aspect is becoming predominant, however, it is
interesting to notice how some systems particularly stress on a specific type of
indicator, such as prices and market trends, food availability, health and
malnutrition. On the other hand, the statistical approach or a complex one
integrating data from various sources are still in use. This is the demonstration of
how heavily the system operation environment does affect methodology.

Indicators and thresholds represent the conceptual content of the information to be
produced. In fact, the indicator is an intermediary step between the input data and
the final information level. Therefore, indicators could either be based on basic data
or on benchmarks, depending on the complexity of the phenomena to be
represented. As regards the thresholds utilised for each indicator, in order to
determine the early warning or risk level, all the same, the variability between one
system and the other and between one application and the other within the same
system is quite important.

5 EWSs and desertification: recent trends and future needs
The existing EWSs utilise environmental and socio-economic data and indicators,
which could be directly employed to assess land degradation or monitor
desertification. The EWSs are deeply focused on food security and they are
approaching other fields of application just occasionally, more for institutional
V

reasons than for technical causes, in order to avoid any possible conflict between
traditional and new stakeholders.

Therefore, only an increasing pressure from users could facilitate an extension of
the field of activity from food security to natural resources management.

Land degradation, being together a cause and an effect of food scarcity, is
generally indirectly monitored by the EWS. Due to its intrinsic correlation with
human and socio-economic factors, desertification could be measured also by
means of the same methodologies utilised by food security.

Nevertheless, some important distinctions must be made. The temporal scale of a
food security EWS is basically conditioned by the rapidity of evolution of the
processes under examination, however, the desertification processes take place in
the medium- and long-term period. The effects of the climatic changes and land
degradation are too slow to be included into a risk analysis such as the crop or
agro-pastoral analysis. As a consequence, the time scale of an EWS for
desertification should be extended over a period of several years in order to see
any remarkable changes in evidence.

Two items seem to be faced with the most serious impact:
§ the micro-scale level analysis of the effects of the populations/environment
dynamics in i) the areas where desertification processes are particularly rapid,
ii) those with population migrations and iii) those where modifications of the
productive systems occur at the same time as remarkable climatic changes;
§ at the regional and global level, the assessment of the status of desertification,
enabling the analysis of the changes occurred in the last decades, for i) the
quantitative assessment of the desertification extension and ii) for the
identification of the degree of vulnerability.

6 Conclusions
As of today, the EWSs are deeply evolving, due to the changes occurred in the
technological environment in which they operate. However, this process should take
into account the present or potential end users, who are facing as well a new
information technology and communication world.

In the near future, some key questions are arising on how to set up a real demand-
driven EWS, rather than a system developed just under the pressure of a
technological push.

Need to develop a common language. The integration between a risk analysis and
a vulnerability analysis, as the structural frame of reference, has become a
generally shared approach. The different meanings of particular terms, i.e.
VI

vulnerability and risk, in different systems is still misleading, making interaction
difficult and causing isolation.

Facilitated access to and transparency of data. As of today, access to baseline
data, in particular, is really neither free nor facilitated, due both to the difficulties in
making the data banks’ network operational and to the idea that data collection
would be the final objective.

Accelerated interaction towards the real partnership. A complex system requires –
especially at this stage –the real willingness to co-operate with a partnership
attitude, vis-à-vis those institutions that might contribute to its development and the
donors/agencies that are asked for the establishment of a political and institutional
‘
enabling environment’ .

Production of a focused information for decision-making. At present, the conceptual
capability of interpreting information is still behind the information production
potential, and the risk arises that an unfocused information will be generated. This
would charge the user with the task of selecting the information, rather than
commanding it.

Users are required to identify the information they need. Users are not a
homogeneous category, as regards both their technical skills and their information
demand. This is certainly a further difficulty facing the EWSs that must decide,
without any active interface, the type of information that is to be provided.

Adequate development of national/sub-national nodes. All the systems under
consideration are operating at regional or sub-regional level, even if they produce
information at the national or local level. How can any national and local EWS be
functionally and institutionally developed, so as to be introduced into the existing
network of systems like those under consideration?

Acceleration of the passage from food security to security. All the early warning
systems under consideration are expanding to new fields of concern, such as
economic planning and management of natural resources. Under this aspect, the
vulnerability analyses are moving towards an improved interaction between
environmental and socio-economic classifications. In this regard, the attention is
being drawn on the concept of ‘ security’ which is based on an organic complex of
,
data and different only as regards the analysis’path.

Technological development should not be considered as a priority. The information
technology is sharply and quickly developing. New generations of satellites are
rapidly becoming operational. Therefore, the EWSs are endowed with theoretically
more and more powerful tools. In this framework, it is of capital importance that -
with respect to these new tools - priority be given to the development of those
applications that would be really suited to the end users.

VII

Acronyms

Centre AGRHYMET Centre Régional de Formation et d'Application en
Agrométéorologie et Hydrologie Opérationnelle
ADDS Africa Data Dissemination Service
ANDI Africa Nutrition Database Initiative
AP3A Alerte Précoce et Prévision des Production Agricoles
AVHRR Advanced Very High Resolution Radiometer
CARPE Central African Regional Program for the Environment
CCD Cold Cloud Duration
CEDARE Centre for Environment and Development for the Arab Region
and Europe
CEO Centre for Earth Observation
CeSIA Centro Studi per l’applicazione dell’Informatica in Agricoltura
CILSS Comité Inter-états pour la Lutte contre la Sécheresse au Sahel
DCW Digital Chart of the World
DEM Digital Elevation Model
ECMWF European Centre for Medium-Range Weather Forecasts
EDC EROS Data Center
EIS Environment Information Systems
EROS Earth Resources Observation Systems
ESA European Spatial Agency
ESRI Environmental Systems Research Institute
EUMETSAT Europe's Meteorological Satellite Organisation
EW Early Warning
FAO Food and Agriculture Organization
FEWS Famine Early Warning System
FIVIMS Food Insecurity and Vulnerability Information and Mapping
Systems
GCOS Global Climate Observing System
GIEWS Global Information and Early Warning System
GIS Geographic Information System
GLCC Global Land Cover Characteristics
GLIS Global Land Information System
GRID Global Resource Information Database
LANDDAAC Land Distributed Active Archive Center
NAP National Action Program
NASA National Aeronautics and Spatial Administration
NDVI Normalized Difference Vegetation Index
NGO Non Governmental Organization
NOAA National Oceanic and Atmospheric Administration
OCDE Organisation de Coopération et de Développement
Economiques
OSS Observatoire du Sahara et du Sahel
PAN Plan d’Action National
VIII

PAR Plan d’ Action Régional
PASR Plan d’ Action Sous Régional
PRGIE Projet Régional pour la Gestion de l’ Information
Environnementale
REWU Regional Early Warning Unit
SAA Satellite Active Archive
SADC Southern Africa Development Community
SAP Système d’ Alerte Précoce
SCOT Services et Conception de systèmes en Observation de la
Terre
SGBD Système de Gestion de la Base de Données
SID System for the circulation of Information on Desertification
SISEI Environmental Information and Monitoring System on the
Internet
SSM Special Sensor Microwave
TIROS Television Infrared Observation Satellite
TOVS TIROS Operational Vertical Sounder
UNEP United Nations Environmental Program
USAID United States Agency for International Development
USGS United States Geological Survey
VAM Vulnerability Analysis and Mapping
WB World Bank
WCMC World Conservation Monitoring Center
WFP World Food Program
WMO World Meteorological Organization
ZAR Zones à Risque

IX

1 Background

The United Nations Convention to Combat Desertification (UNCCD)1 was signed in Paris
on 17 June 1994 and came into force on 26 December 1996. It provides the innovative
framework for the sustainable development, in arid, semi-arid, dry sub-humid areas, of an
appropriate implementation mechanism to combat desertification at the global, regional,
sub-regional and national levels.
At its second session, the Conference of the Parties (COP), which is the Convention's
supreme body, held in Dakar from 30 November to 11 December 1998 and adopted - inter
alia- the following decisions:
- to invite Governments to initiate testing the application and the impact indicators as
well as the practice of using those indicators in national reporting to the third session of
the COP for the Affected African country Parties;
- to decide that the priority issue to be addressed in depth by the Committee on Science
and Technology (CST) at its third session shall be early warning systems in the
broadest sense. Moreover, in order to support the action of CST in regard to this topic,
the Parties are invited to submit contributions reporting on the existing experiences of
early warning systems, while the specialised Institutions acting in this field are required
to facilitate the preparation of the third session.

At present, many early warning systems are not addressing environmental issues, in
particular desertification, both in terms of indicators and of factors affecting food security.
At the same time, the systems dealing with the monitoring of natural resources have given
priority to the environmental aspects, leaving aside the human factor, as the affected and
affecting element of the status of desertification.

This initiative - which is to be considered in the above context - aims at providing a better
knowledge of the dynamics existing between food security and desertification, as well as to
support the integration of the desertification dimension in the existing early warning
systems, so as to:
- support the action of the Committee on Science and Technology by producing an
evaluation on the problems arising when integrating the environmental dimension -
desertification in particular - in the early warning systems.

1
The Convention to Combat Desertification was negotiated under the auspices of the
United Nations. In June 1992, the United Nations Conference on Environment and
Development (UNCED -- also known as the Rio Earth Summit) recommended that the
United Nations General Assembly establish an Intergovernmental Negotiating Committee
(INCD) to prepare a convention to combat desertification in those countries experiencing
serious drought and/or desertification, particularly in Africa. The Committee was
established in early 1993. It held five preparatory sessions before adopting the Convention
on 17 June 1994 in Paris.

1

- create an “enabling environment” among the systems currently under implementation
in the fields of early warning and/or environmental monitoring over Africa, bearing an
environmental dimension, in particular as regards desertification.

This report is a main part of the initiative sponsored by the Italian co-operation, which
include a four-day workshop to be held in Niamey, Niger, from 25 to 28 October 1999. It
provides a framework of comparison among the various existing systems operating in the
fields of early warning, monitoring and environmental data dissemination and production,
that are accessible through the Internet. The quantity of information available as of today is
a unique wealth for the implementation of the UNCCD and it can be promptly employed
thanks to a closer collaboration among the projects/institutions called to produce a demand-
driven information.

This report was prepared by the CeSIA - Accademia dei Georgofili on the basis of the
arrangements between the CCD Secretariat and the Italian Co-operation.

2 Introduction

Late in the ‘70s, as a consequence of the dramatic drought occurred in West and East Africa,
the famine struck millions of people. The affected areas’ administrations and the
international community were faced with the need to provide the appropriate tools to
facilitate mobilising measures aimed at mitigating the impact of recurrent droughts.

In those years, for example, the ORSEC Plan, proposed by the French Co-operation first,
and subsequently resumed by the European Community, aimed at ensuring adequate food
supply to the Sahel populations in case of crisis.

The plan was characterised by a “demand for information” clearly identified in terms of
users, contents, forms and timing for utilisation, on the basis of a detailed analysis of the
factors affecting the Sahelian populations’ food security in case of repeated drought in the
forthcoming years.

While the analysis recognised the climatic factor to be the key-element in the repetition of
the crisis condition, however, famine prevention was strictly depending on the possibility to
make available – by the end of July - a reliable information on expected generalised famine
conditions. In fact, the effective distribution of food aids was a time-demanding operation,
due to: i) the decision-making processes for the supply of assistance by the international co-
operations, and ii) the harbour facilities and the primary transport systems in West Africa.

In the most vulnerable areas (350-600 mm of rainfall) the cropping season start at the
beginning of July, therefore the relevant information could not rely upon agricultural
statistics but only on yield forecasting.

2

The plan was a rare example of a demand-driven system, rather than a technological push.
Unfortunately, due to the limits of technologies and knowledge available at that time, it was
impossible to implement the whole system and therefore the initiative aborted.

The early warning systems that were conceived and implemented in that period can be
considered the ancestors of the systems in use today. In fact, from a methodological point of
view, they aimed at forecasting the establishment of a risk situation on two scales: i) the
geographic area and ii) the population involved.

All along the ‘ 70s, famines were mainly considered as events deriving from particularly
severe dry years, the first early warning systems were, therefore, based on a close
correlation between food insecurity and the meteorological trends of the year.

On the other hand, the agro-meteorological evaluation models of crop yields, were not only
the sole forecasting instruments available at that time but also the only ones enabling the
supply of a homogeneous information on vast areas. However, the information produced by
those instruments resulted insufficient for the processing of the information regarding the
years that are not particularly dry - when only small areas are struck and the establishment
of crisis conditions is mainly due to the characteristics of the productive/socio-economic
systems.

Since then, as a consequence of both the great results expected from the early warning
systems - due to their theoretical potentiality as well as to the remarkable financial
investments involved - and the unimpressive results obtained, a passionate debate was
generated at the international level to find the most appropriate solution.

Some misunderstanding arose by the use of the same terminology to characterise early
warning systems aiming at different objectives or focussing on different kinds of
phenomena. This is demonstrated by the vastness of the bibliography existing on what
should be an early warning system, and the one concerning the experience gained by the
existing early warning systems in the latest years.

These unsatisfactory results were the consequence of structural constraints such as:

- the capability of the simultaneous management of data originated by satellites and
agro-meteorological models for analysis was available only in specialised centres
endowed with the adequate information equipment, where, similarly, a remarkable
specialisation level was required;
- the specialisation level required to integrate socio-economic and bio-physical data
processing by very complex techniques;
- the lack of techniques and methodologies appropriate for the processing of information
at the territorial level, especially for those data that could not be given a homogeneous
coverage;

3

- the transfer process of data to and from the specialised centre for the production of
information required a time length inappropriate to the intervention that would be
required on the phenomena observed, therefore making it useless;
- the final information was neither synthetic nor appropriate for decision-making;
- the appearance on the scene of new food insecurity causes linked to socio-economic
and political factors.

The latest solutions were focused on three main axes:
- the creation of national systems based on the channelling of all thematic information
coming from the different national services towards a common clearing-house
mechanism, in order to have at least a qualitative view of the evolution of the crops and
of the food risk conditions arising; this approach enabled to orderly process the
information incoming to the centre, but could ensure neither a homogeneous quality
nor a medium-term forecasting;
- the acknowledgement of the nutritional aspects and of those linked to the access to
goods as indicators more appropriate for the monitoring of food security; the
impossibility of managing the data available on a homogeneous geographic basis has
never completely enabled this approach to become operative;
- emergency being transformed from an exceptional into a regular event, by giving
priority to a more efficient organisation of aid to be constantly operational with respect
to their mobilisation only in case of unforeseen emergency. This approach enabled an
improved response to the multiplying crises, mostly deriving from political reasons,
that have been faced - especially in the last years - but demanded very large financial
resources that had to be taken away from the development operations, thus generating
an impact on the populations’food security on the long-term.

It was only in the mid ‘ 90s, when the technological evolution in telecommunications (the
Internet) and information technology (hardware and software) took place in parallel, that the
conceptual revision of the structural constraints of the technological type enabled the
parallel experimentation of new operational approaches.

In fact, while the baseline data and the data collection systems are the same as twenty years
ago, in the course of the last five years, the capability of processing and managing data in
order to produce information and facilitate access to information, has gone through a deep
revolution. In fact, the same information that was earlier:
§ punctual, is now spatialised
§ qualitative, is now quantitative
§ simple and sectoral, is now complex and synthetic
§ limitedly disseminated, is now accessible to all
§ produced by few specialised centres, is now generated in a decentralised mode and with
no need for any particular specialisation

Systems that are now in a position to provide forecasts of different crisis-risk levels on a
micro scale according to geographic area and involved population (risk villages, target
4

groups, etc.) are now getting through the experimental phase and reaching the operational
stage.

Based on the above elements, the Early Warning Systems are still evolving, thanks to the
development of the vulnerability mapping systems, towards the integrated management of
the structural vulnerability analysis. The last is meant as the capability of a population, a
village, a social group to face a negative event, with the risk forecast, that is the possibility
that a negative event occur in a given time.

At present, the future scenario seems particularly favourable to early warning, as it is
characterised by:
§ an increasingly accessible and timely information;
§ the development of data integration techniques aimed at producing an immediate,
useful and diversified information, according to the needs of the different end users.

An analysis of the Early Warning Systems must be necessarily based on those future
development aspects that are foreseeable from both the technological point of view and from
the point of view of users as well as of their needs. In this evolution framework, they keep
their purpose to represent an instrument both for rapid intervention and for the assessment of
the causes of fragility of a given agro-food/environmental system, even if they are radically
changing their characteristics.

It is now time to speak in terms of families of specialised systems, aiming at supplying the
information that is adequate to a specific user. In this way, each system is no longer aimed at
being an autonomous and self-sufficient reality, but the node of a network of systems – or
‘entities’– which will facilitate its operations.

In fact, a complex and global system is being developed, formed by ‘ entities’ producing and
distributing processed data, those which are more immediately involved in early warning
and those which create a favourable environment for the circulation of information. This is
the reason why this analysis focuses on the group of ‘ entities’ that have chosen the Internet
as the enabling environment in which they can operate in order to ensure - at the same time -
both the free access and the rapid exchange of information.

3 The present scenario

An Early Warning System is based on three main components, namely:
1. the collection of data,
2. the processing of data and the production of information, and
3. the dissemination of information.

These three elements can be found all concentrated in the same body or even separate one
from another. Even though the present trend is demanding a more flexible view, we are still
defining the Early Warning System as the one encompassing all the three elements. On the
other hand, also the other possible ‘entities’ must be considered, that is the information

5

dissemination systems and the producers/suppliers of the basic datum, which we shall call
data dissemination services.

Graph 1Early Warning System

Data Dissemination
Service

Data processing and
information production
System

Information Dissemination
System

End users

On the basis of this classification, a comparative analysis of the systems operating in the
Internet has been undertaken in the last months. As a result, not so many Early Warning
Systems can be considered as such in the strict sense of this expression. While the
Information Dissemination Systems are still less numerous, there is a definitely larger
number of Data Dissemination Systems. All the three definite categories will be considered,
namely:
1. Early Warning Systems, in the strict sense of the word;
2. Information dissemination systems;
3. Data dissemination systems.

A compared analysis of these Systems will be presented in the following sections. The bulk
of information produced is the one supplied by the Systems themselves through the
respective Internet sites. The most detailed analysis was carried out regarding the first class
of Systems, by examining both the institutional and the methodological aspects. The systems
of information dissemination and the data dissemination systems were considered mostly in
view of the early warning application to desertification.

6

Table 1 List of Systems
Name URL
Early Warning Systems
Agrhymet Alerte Précoce et Prévision des Production http://www.iata.fi.cnr.it/ap3a/ap3a.htm
Agricoles (AP3A) project
USAID's Famine Early Warning System (FEWS) http://www.info.usaid.gov/fews/
SADC Food Security Programme (/REWU) http://www.zimbabwe.net/sadc-fanr/
FAO Global Information and Early Warning System http://www.fao.org/giews/english/giewse.
(GIEWS) on Food and Agriculture htm
FAO Food Insecurity and Vulnerability Information http://www.fivims.net/
and Mapping Systems (FIVIMS)
WFP Vulnerability Analysis and Mapping (VAM) http://www.wfp.it/vam/vamhome.htm
Information Dissemination Systems
WB PRGIE
WB Environment Information Systems (EIS) in Sub- http://www.grida.no/eis-ssa/index.htm
Saharan Africa
OSS System for the circulation of Information on http://www.bondy.orstom.fr/sid-oss/
Desertification (SID) / Environmental Information and
Monitoring System on the Internet (SISEI)
Scot Conseil and Medias-France Desertification Data http://www.scot-
and Information System (D-DIS) sa.com/scotnew/frame_f.htm
CEO - Desertification Information Network http://www.wcmc.org.uk/dynamic/desert/
Data Dissemination Syatems
ESA IONIA http://shark1.esrin.esa.it/
ESRI Digital Chart of the World (DCW) http://www.esri.com
EUMETSAT http://www.eumetsat.de/en
FAO-AFRICOVER http://www.africover.org/
FAOSTAT http://apps.fao.org/
NOAA Satellite Active Archive (SAA) http://www.saa.noaa.gov
http://daac.gsfc.nasa.gov/CAMPAIGN_D
NOAA/NASA Pathfinder AVHRR Land FTP
OCS/FTP_SITE/readmes/pal.html#100
PENN STATE UNIVERSITY, DCW Data Server http://www.maproom.psu.edu/dcw/
UN Africa Nutrition Database Initiative (ANDI) http://www.africanutrition.net/
UNEP Global Resource Information Database (GRID) http://grid2.cr.usgs.gov/
http://carpe.gecp.virginia.edu/partners/gsf
USAID CARPE
c-umd/UMD/gisthemes.html
USGS Africa Data Dissemination Center (ADDS) http://edcintl.cr.usgs.gov/adds/adds.html
USGS Earth Resources Observation Systems (EROS) http://edcwww.cr.usgs.gov/dsprod/prod.h
Data Center (EDC) tml
http://edcwww.cr.usgs.gov/landdaac/1K
USGS LANDDAAC
M/comp10d.html
USGS Global Land Information System (GLIS) http://edcwww.cr.usgs.gov/webglis
http://edcwww.cr.usgs.gov/landdaac/glcc/
USGS Global Land Cover Characteristics
glcc.html

7

Centre for Environment and Development for the Arab
http://www.cedare.org.eg/gis
Region and Europe (CEDARE) GIS
WMO Global Climate Observing System (GCOS) http://www.wmo.ch/
European Centre for Medium-Range Weather Forecasts
http://www.ecmwf.int/charts/charts.html
(ECMWF)
World Conservation Monitoring Center (WCMC)- http://www.wcmc.org.uk/forest/data/cdro
Forest Conservation m2/index.html

4 Early Warning Systems

The analysis for the six Early Warning Systems already in operation is carried out on three
different levels:
§ Objectives, products and users.
§ Methodological aspect, scale of application and indicators employed, and field of
application.
§ Circulation of information mechanism

4.1. Objectives, Products and Users

4.1.1. Objectives
The final objective of any Early Warning Systems is to provide the decision-makers with the
necessary and timely information on the present food situation in the relevant areas, and the
forecast for the end-of-season. As far as objectives and products are concerned, every single
system is characterised by differences depending on the geographic area of application.
These differences are often due to the availability of financial resources, the availability of
data in the national setting, and to the specific agency requirements for the structure and the
contents of outputs.

The objectives of the various Early Warning Systems can be broadly defined on the basis of
two different scale levels, that is national/regional and global. At the national/regional level,
the first objective identified by GIEWS in the ‘ 70s is the assessment of food supply and
demand, that is the assessment of the food situation. This is the historical framework of all
the traditional Food Security and Early Warning Systems. Differences exist as regards the
application scale, in fact, while GIEWS follows a global approach, when the assessment is
carried out on all the countries, other systems focus on specific countries or regions –
intended as a group of countries – as, for example, the SADC region. Some Systems, like
AP3A are centred, as of today, on the concept of risk, some other Systems, like WFP-VAM
and FEWS, focus on the integration between risk and vulnerability.

All the Systems, more or less, make a distinction between a statistical analysis of the
situation, which is defined ‘structural analysis’ and a dynamic analysis, which is defined
‘conjunctural’ The structural analysis does not take into account the time dimension,
.
therefore, it cannot represent the progressing development of a phenomenon, while the
8

conjunctural analysis, in order to be significative, needs a structural frame of reference. As a
consequence, it is clear that the two analyses are complementary and integrate with each
other. In this framework are to be included both the monitoring of the agricultural season
and the changes that can lead to risk situations (FEWS).

For Systems like AP3A and SADC, the need to determine risk areas leads the analysis to a
geographic connotation; the territorial units utilised can either be defined as administrative
units at various levels, or homogeneous units according to certain characteristics. On the
other hand, both FEWS at first, and SADC later, introduce also the concept of human groups
to identify populations in the place of areas, which still represents a criterion of great
complexity when shifting from the local to the sub-national and national analyses.
Table 2 Objectives
Systems Objectives
WFP-VAM Provide an analysis of the baseline data and current vulnerability patterns together
with their causality
Link the analysis of vulnerability with specific operational programming decisions
and problems in the country strategy
Strengthen on-going contingency planning exercises and other disaster preparedness
measures
Integrate VAM analytical techniques into country structure
SADC Provide member states and international community with advance information on
food security prospects in the region through assessment of expected food production,
food supplies and requirements and definition of food insecurity areas and population
FAO- At the national level:
FIVIMS Improve policy formulation
Improve programme management
More effective design and targeting of interventions
More effective inter-sectoral and inter-institutional dialogue
At the global level:
link relevant data from existing internationally-held databases through a commonly-
accessible data dissemination system, available in the public domain
FAO- Monitoring food supply and demand in all countries of the world on a continuous
GIEWS basis
Providing information on global production, stocks, trade, export prices
Developing new approaches to early warning
Sending rapid food supply and demand evaluation missions to the affected countries
Report to the international community its regular publications
USAID- Improve understanding of the basic causes and circumstances of famine
FEWS Detect changes that create serious famine risks
Vulnerability assessment
Determine appropriate famine mitigation and prevention strategies
AGRHYME Develop a system addressed to help SAP in decision-making
T-AP3A Develop methodologies for the definition of areas at structural and current risk, in
agriculture and breeding fields.
Provide the International Community with useful information to define investment
policies

9

Another step forward is to be considered the aim of permanently integrating these analysis
procedures with the methodology utilised in the mechanism of each country’ policy s
formulation. What above reflects in the VAM ‘ vulnerability analysis integration with each
country’ strategies of planning policies’ or by ‘
s strengthening the decision-making process’
in AP3A or by ‘ improving the capability to formulate the national policies, thus allowing a
better-focussed definition for the interventions of FIVIMS. The latter also proposes the
objective of improving relationships and dialogue among the different sectors and
institutions operating in the field of food security.

At the global level, the objectives are submitted to the attention of the international
community, donors and international agencies. The more generic objective is to provide the
international community with the useful information, thus enabling the formulation of
investment policies, in order to improve food security programmes and reach an in-depth
comprehension of the phenomenon of “hunger”. Moreover, some other Systems, such as
FIVIMS, AP3A and FEWS have defined their objective of producing a data dissemination
system - to be common property and knowledge - capable of integrating the flows of data
coming from different sources and databases. Two additional general objectives of GIEWS
are i) to provide the international community with regular publications on food security and
ii) to develop new approaches for the Early Warning.

4.1.2 Products
It is difficult to give a precise and univocal definition of the products provided by the
different systems. In fact, the products are very much case-specific ones. The smaller, AP3A
or SADC, systems, are characterised by a series of standard products, applied to the
different countries in a more or less systematic way.

On the other hand, WFP or FEWS depending on the situation - of data inflow and goals to
be reached - produce different types of information. The most complete and complex
products are WFP and FEWS: they include the analysis of conjunctural risk of both
structural and conjunctural vulnerability. As regards the conjunctural risk analysis, it is
possible to find the products of the AP3A models, including a very precise indication of risk
zone according to the different factors.
All the products of the GIEWS, FEWS and SADC’ statistical analysis of agricultural
s
productions and market trends are placed on a different information level.

For all these systems, the methodology and procedure development is “ se” a product to
per
be diffused, while for FIVIMS this is the most important one.
Moreover, the differences among the various types of products provided should be
considered also with regard to the medium and format. For example, some systems, like
GIEWS or WFP, or AP3A, can provide users and the public with bulletins as well as with
digital information, information planes, risk charts, etc. In some cases it is even possible to
accede to special databases containing the very baseline data utilised in processing (ADDS
of FEWS, SGBD of AP3A and FAOSTAT of FAO).
10

Table 3 Products

Products AP3A FIVIMS GIEWS SADC FEWS VAM

Food insecure areas and population
Food insecurity assessment bulletins
Crops situations reports and bulletins
Food supply/demand situation bulletin
Special Reports and Alerts
Methodologies
Crop risk index current and baseline
Coping strategies/assets index
Market access index
Vulnerability index: current and baseline

Products format

Bulletins
Regular publications
Assessment maps on digital format
Informative layers on digital format
FAO ADDS
Databases on gross or elaborated data SGBD
STAT EDC
Methodology description
Software application

4.1.3 Users
The end users of the early warning system should be the monitored populations (target
groups). However, in general, information is not directly reaching these users and it is
filtered through the national/local institutions. In the case of early warning for food security,
information is mainly aimed at providing the most objective basis on which the user
working on it at any institutional level, can make the best decisions on what are the
necessary actions to be taken.

At present, especially in the field of food security, the decision-making process, such as for
example “a given area could be affected by food shortage and it is necessary to send food
aid” normally involves various actors at the local, national and international level:
- local administrators, who should properly inform at the central level those
responsible for taking action,

11

- national services, who should provide at the central level analysis and forecast on
possible risk areas,
- national focal point, which should sort a synthetic information from all the
available data and specific information
- political figure in charge, who should make the decision and approve the
intervention
- international partners, who should support the intervention by mobilising the
necessary resources

Moreover other actors, such as NGOs or the international community, could either intervene
independently in the affected areas or could co-ordinate their intervention with the national
authorities.

It is clear that an early warning system - as it is the case for Agrhymet-AP3A and the
various CILSS countries - is not addressing the information to all the involved actors but has
preferred users, in particular the national services, the national focal point able to process the
incoming information.

This type of users are not, therefore, end users and disseminators of information, they are
entities that, in turn, should produce further information.

For most Early Warning Systems, another important user is the international community
including NGOs as well as the headquarters of internationals organisations and aid agencies.
Obviously, the type of information utilised by these users is aimed at assessing country
priorities in emergency interventions and for long-term action, in particular, for the
definition of strategies and policies and the identification of target groups and areas of
intervention. Finally a completely different user is the scientific community interested in
receiving data and information to support the development of new methodologies or
techniques.

Graph 2 Users

Users
Other Food
Security projects
NGO
19%
8%
Public
National institutions
12%
8%

National
governments
22%
International
Local Organisations
administrations Local institutions 23%
4% 4%

12

In principle, depending on the type of users involved, the types of needs will be different
and obviously also the frequency of the information request will change. For instance, those
who must decide need very synthetic, reliable and up-to-date information, therefore, at the
national or local level, the information needed will be of the aggregate type, easy to
understand, explanatory of the situation in the most general way. On the contrary, those who
must process information, to obtain a more synthetic product, need - even if in a schematic
format - a view of the market situation, food demand and supply, trend of the crop year and,
on the other hand, a thematic cartography of food vulnerability, in risk zones, insufficient
production areas, etc.

As regards the international community, the requested products will be the overview of the
food security situation, vulnerability classification, possible risk areas, etc.

4.2 Methodology

4.2.1 Type of system
According to the most modern conception, the Early Warning systems are based on a very
extensive multidisciplinary analysis. The socio-economic aspect is evidently predominant,
however, it is interesting to notice how some systems particularly stress on prices and
market trends (FEWS), food availability (WFP-VAM), health and malnutrition (FIVIMS,
FEWS). On the other hand, a more statistical, agricultural and food approach is the one
adopted by GIEWS, while AP3A extends its range of action from agro-meteorology to
livestock, also integrating them with socio-economic baseline information. The utilisation
of remote sensing for the provision of meteorological information, also integrated by ground
data, vegetation and land cover maps is the common denominator.

As regards the scale of information utilisation, as it will be clear further on, some systems,
like AP3A, WFP-VAM or SADC, are different, as they also refer to third-level sub-national
administrative units, while some others, like GIEWS, are only concerned with higher levels.

On the one side, from the point of view of information quantity, GIEWS is the most
extensive one. It may suffice to know that it constantly follows the demand and supply trend
of the world cereal market, by elaborating analyses on the production, stocks, export price
levels, the world trade and aid. This activity of GIEWS, at the world level, is based on an
very large data collection network and supported by evaluation missions in the countries
involved. The management of this type of system is faced with the double problem of data
accessibility and reliability.

However, FEWS applies the most complete and complex methodology, as it follows a
multidisciplinary approach and is scientifically supported by USGS and EROS. Starting
from both the remote sensing and agro-meteorological data, FEWS finds the basis necessary
to identify the areas exposed to potential famine. At the same time, by exploiting the data
concerning food availability and accessibility, it can assess the vulnerability of both
geographic areas and human groups. Household food insecurity is evaluated at the lowest
possible level of desegregation (typically the third administrative unit), incorporating
13

information on the different socio-economic groups within each area to the greatest possible
extent.

Graph 3 Data, elaboration and information flow

DATA
Gross data (rainfall, Partially analyzed data (NOAA and
temperature, productions… ) METEOSAT images Indexes… )

ELABORATIONS

Data geographic distribution, spatialisation and missing data

Data analysis, integration of GIS, Data Bases and models

System flexibility, scale changing procedures

Data crossing, indicators individuation, structural and impact indicators,
indicators extracting methodology; quantitative and qualitative data;
algebraic synthesis, overlaying of different informative layers.

Risk and vulnerability: territory structural zoning on the basis of vulnerability
analysis and risk evaluation (methodology for risk identification or
unfavorable events probability and their impact evaluation); utilization of a
scale coherent with the methodologies applied in planning; homogeneity
between the classification methodologies used for the different countries

PRODUCED INFORMATION
Vulnerability indicators Impact indicators Base Maps

Vulnerability maps Risk maps

Early Warning and catastrophic events prevention

FEEDBACK INFORMATION

Users needs Particular elaboration requests Implementation indicators

The approach of VAM is the one which differs from the others. The VAM methodology
baseline is given by the integration between the assessment of food accessibility (or the
population’ capability of reaction in case of unfavourable events) and the analysis of risks.
s
The latter is an integrated analysis in relation with some particular factors such as :
14

Flood/Drought, Vegetative Conditions (NDVI Analysis), Pest Infestation, Market Price
Changes and Health Conditions. The assessment of vulnerability is based on the data
obtained by the integration between food accessibility and risk.

The AP3A methodology is the one which differs from the others, in fact, instead of being
centred on the most traditional economic aspects (prices, markets, etc.), this approach pays
great attention to the agro-meteorological and agro-pastoral analyses. Agricultural
production is the factor that mostly determines food availability: in the Sahel region this
factor is based on rainfed crops and is mostly destined to self-consumption. Therefore, the
so-called food risks zones are those where the rainfed cereal production is insufficient. The
agro-meteorological aspect is, anyway, integrated with the socio-economic aspects,
represented by basic information on the agricultural and pastoral production, data on
population etc.

On the contrary, FIVIMS differs from all the other systems because, more than an Early
Warning system, it is a network of systems, a connection through which food insecurity
information is gathered, analysed and collected. At the global level, FIVIMS, somehow,
follows the GIEWS approach, still broadening the nutritional aspects. At the country level,
the FIVIMS approach serves as support to the development of national Early Warning
systems. Basically, FIVIMS is not so much an Early Warning in the strict sense of the
expression, but it is a tool for the collection of information, study and phenomenon location.

The different systems’approach to data analysis is based on two main lines. The first line of
approach is of the statistical type, which is markedly utilised in the socio-economic data
analysis as well as in the analysis of all the data that have not been geographically
distributed. The statistical approach is based, in general, on the analysis of principal
components.

The second line of approach regards the geographic information systems, which include
both the analysis and the integration of data that are different at the geographic level. This
approach is generally utilised for environmental and land data, namely, all the information
regarding the territorial aspects. These two largely general approaches integrate each other
into a more complex analysis, which reports both the data and the statistical analyses
according to the geographic aspects.

4.2.2 Scale of application
The geographical scale factor does not characterise each system; inside each system there
are different applications and different scales. Anyway, in general, one can say that some
systems are conceived to work at the global level, as for example, GIEWS or WFP-VAM,
some others at the continent level, like FEWS, while some of them, like AP3A and SADC
work at the regional-national level. Therefore, the trend is from the global to the local level,
still with particular emphasis on the national level.

The basic territorial unit is generally an administrative one, from the first to the third sub-
national level. As an alternative, the basic territorial unit can be other than the administrative
15

one, as the case is for AP3A and all those projects which utilise agro-climatic,
environmental and socio-economic zoning. A land classification according to one or more
parameters is obtained by using either one of the two approaches or both of them
simultaneously.

On the other hand, the analysis is not always carried out at a precisely geographic level. In
fact, according to FEWS, the assessment of vulnerability does concern both some
geographic regions and some human groups. Within one area, there may be several socio-
economic groups, each one with its own distinct way of accessing food (small farmers,
pastoralists, fisher-folk, refugees, female-headed households, urban dwellers, etc.).
Depending on the different production strategies - that is also revenue - the various human
groups are classified into socio-economic groups; it is on these groups that FEWS developed
its methodology for the assessment of vulnerability. A geographic region for FEWS is a
dynamic territorial system presenting homogenous internal characteristics, referred, in this
case, to food security.

4.2.3 Indicators and thresholds
Indicators are defined by the OECD2 as a value calculated starting from a group of
parameters providing information on a phenomenon or its condition. The indicator is aiming
at a certain objective and addressed to a certain type of users. It reflects a situation and can
help decision-making in that particular context. The indicator can either be the measure for a
quantitative evaluation or the qualitative element to describe a situation.
The application of thresholds to quantity indicators is utilised for the representation of the
various levels of a phenomenon intensity. The choice of thresholds, which are to be set up
for each indicator, is very important as it is aimed at explaining how serious the situation is.

Since the ‘70s, the main indicators utilised for the Early Warning systems were mostly those
related to the meteorological season and crops growth; later on, in the ‘ 80s they oriented
towards remote sensing and nutritional data, to reach the socio-economic aspects in the ‘ 90s.
Indicators and thresholds, being rather specifically established for each application - as it is
the case for the data utilised and the products obtained - represent the conceptual content of
the information to be produced. In fact, the indicator is an intermediary step between the
input data and the final information level. Therefore, indicators could be based on basic
data or on indicators, depending on the complexity of the phenomena to be represented in
the information.

Obviously, with more specific systems, like AP3A or SADC, they can basically utilise the
same set of indicators for the different countries of application. In fact, their field of action
is at the regional level, while for some projects like GIEWS or WFP-VAM- being the field
of action far more extended - it requires a wider range of indicators. Anyway the indicators
utilised basically reflect the approach followed by the various systems.

2
OCDE/GD(93)179, Monographies sur l’ environnement, N° 83: Corps central d’indicateurs de l’OCDE
pour les examens des performances environnementales, Paris 1993.
16

Table 4 Indicators
Indicators AP3A FIVIMS GIEWS SADC FEWS VAM

Food crop performance
Crop conditions
Crop production forecast
Marketing and price information
Food supply/demand
Health conditions
Food crops and Shortages
Food supply
Food comsumption
Crop areas
Pests
Food balance
Vegetation front
CCD
NDVI
Biomass
Seeding risk areas
Expected season length
Estimated seeded areas
Estimated seeding date
Vegetation cover
Agro-ecological zones
Crop use intensity
Variation coefficient of agricultural
production
Cash crop production area
Coping strategies
Average cost to travel to nearest market
Livestock production
Population density
Access to water
Children education
Rainfall

17

Therefore, we can notice that the AP3A indicators are basically agro-meteorological ones,
while the GIEWS indicators follow the approach of agricultural and food statistics, and the
WFP or the SADC indicators are mostly agro-economic ones. The meteorological
indicators are vastly used by all the projects and are produced by remote sensing
information, which is generally integrated or gauged on ground data.

The agro-meteorological indicators are to be considered on a different level. This type of
indicators synthesise and aggregate information related to both the climate and the cropping
characteristics of soils. The utilisation of these indicators is in close connection with the
agro-meteorological follow-up of crops, which has witnessed the development of a vast
variety of simulation and/or forecasting models, which are very often used in the Early
Warning systems.

The socio-economic indicators take into consideration the nutritional conditions of
populations, the market macro-economic aspects and the coping capacity, that is the
population’ capacity of reaction when faced with a certain unfavourable event. As regards
s
these data, collection and, therefore, monitoring are easier for the market data, in fact,
GIEWS produce a global coverage. On the contrary, the coping capacity assessment is a
difficult operation, as the various human groups often provide very particular and complex
responses. As a consequence, an effective coping capacity assessment is made only for
some very delimited applications (VAM).

As regards the thresholds utilised for each indicator, in order to determine the early warning
or risk level, all the same, the variability between one project and the other and between one
application and the other within the same project is quite important. The subject is even
more delicate as regards thresholds, in fact, also for countries that are very similar to each
other - or even within the same country - there can be a difference in “sensitivity”. Here we
are back to the field of the vulnerability analysis, that is the capacity or incapacity of the -
natural or human - environment to react when faced with a negative event, therefore, it has
to be based on complex indicators, that are often based on socio-economic and bio-physical
indicators.

4.2.4 Geographical distribution
The majority of Early Warning Systems are targeted on Africa, which gathers more than
60% of the systems. GIEWS, FIVIMS and WFP-VAM are also involved in other countries,
mainly Asia and Central and Latin America.

Africa is to be considered as divided into three main areas or regions, Western Africa,
Southern Africa and Eastern Africa. GIEWS, WFP-VAM and FEWS cover all the three
sub-regions, while SADC covers only Southern Africa and AP3A part of Western Africa
(CILSS countries).

18

Table 5 Geographical distribution of Early Warning Systems
Countries AP3A FIVIMS GIEWS SADC FEWS VAM

Afghanistan
Algeria
Angola
Bahrain
Bangladesh
Benin
Bhutan
Bolivia
Botswana
Brazil
Brunei Darussalam
Burkina Faso
Burundi
Cambodia
Cameroon
Cape Verde
Central African Republic
China
Comoros
Congo
Costa Rica
Cuba
Cyprus
Djibouti
Dominica
East Timor
Ecuador
Egypt
Equatorial Guinea
Eritrea
Ethiopia
Gabon
Gambia
Ghana
Guatemala
Guinea
Guinea-Bissau
19


India
Indonesia
Iran
Iraq
Israel
Ivory Coast
Japan
Jordan
Kenya
Korea
Korea, Democratic Republic
Kuwait
Laos
Lebanon
Lesotho
Liberia
Libyan Arab Jamahiriya
Macao
Madagascar
Malawi
Malaysia
Maldives
Mali
Mauritania
Mauritius
Mayotte
Mongolia
Morocco
Mozambique
Myanmar
Namibia
Niger
Nigeria

20


Pakistan
Peru
Reunion
Rwanda
Sao Tome and Principe
Senegal
Seychelles
Sierra Leone
Somalia
South Africa
Sri Lanka
Sudan
Swaziland
Tanzania
Tchad
Togo
Tunisia
Turkey
Uganda
Venezuela
Zambia
Zimbabwe

Graph 4 Systems geographical distribution
Center-South
America
6%
Asia
26%

Africa
68%
21

4.3 Mechanism of access to information

Early warning makes no sense if it is not followed by any action; action takes place only if
the information is timely disseminated. Therefore, the information dissemination
component of the system plays a very important role in Early Warning System. Most of
them are based on a more or less extensive network of data collection as well as on a
dissemination network and a comparison between the two is not always possible.

As a matter of fact, information is not always efficaciously disseminated yet, nevertheless,
this is a vital and crucial component for every Early Warning System. The main reason is
that traditional techniques are more and more out-of-date, while new technologies - enabling
a timeliness which would not have been possible a few years ago - have not reached a
sufficient level of diffusion among users.

One very clear example is GIEWS, which gathers information about almost all the world
countries through an enormous data collection network, which is not anyway gauged on the
one which disseminates the results. Indeed, it should be stressed that an Early Warning
System should be conceived in order to facilitate the access to the produced information
based on a technology going beyond the national and administrative borders: the Internet
has become the great instrument.

All the Early Warning Systems have their own web site, where bulletins, reports and
documentation are easily accessible. However, the main numeric products of data
processing or any indicators, such as maps, are not always disseminated through the
Internet. It must be pointed out that, even those which provide free access to thematic,
vulnerability and risk maps (GIEWS, AP3A, FEWS and WFP-VAM) do not ordinarily
diffuse the same maps in a format enabling their utilisation by a GIS.

On the contrary, it is possible to find database online, which give the possibility of
downloading of both baseline data and partially processed data.

Bulletins are still the most traditional output to be circulated by an Early Warning System.
For instance all the Early Warning Systems have a regular bulletin online all along the year
or during the crop season. Bulletins are often supported by special reports and periodicals or
discontinuous publications on thematic evaluation or assessments or methodology.

5 Information Dissemination Systems

The information dissemination systems mainly focus on environment themes and, among
them, desertification plays the most important role. As of today the more developed systems
are:
1. WB-Projet Régional pour la Gestion de l’ Information Environnementale (PRGIE),
2. WB Environment Information Systems (EIS) in Sub-Saharan Africa

22

3. OSS System for the circulation of Information on Desertification (SID) /
Environmental Information and Monitoring System on the Internet (SISEI),
4. Scot Conseil and Medias-France Desertification Data and Information System (D-DIS)
5. CEO - Desertification Information Network,
6. CEO - EWSE Information Exchange for Earth Observation

As a general rule, all these systems do not directly produce any information but are aimed at
disseminating information or facilitating access to the information produced by others like
the Early Warning Systems or Data Dissemination Systems.

The Internet is no doubt the favourite environment even though some systems, PRGIE and
EIS, are building parallel networks not accessible from outside, with the objective of
creating an enabling environment among the users.

CEO - Desertification Information Network focuses on desertification, however it is not
completely developed yet. The EWSE, which is connected with CEO, is an on-line
information service for the Earth Observation community. Even if, as a matter of fact, the
EWSE is rather a list of Internet sites than a real system for disseminating information. The
D-DIS of Scot Conseil is still being tested, therefore, at present information is only available
on CD-ROM.

In fact, it is only OSS-SID that is making information available on the Internet. The
information available concern territorial analysis through biophysical indicators and baseline
data. The most complete and aggregate information that OSS-SID is able to supply is a
classification of evapo-transpiration and of vegetation.

The objectives of OSS-SID are to:
1. allow, at different scales, to accede and disseminate selected information, validated and
made available under such forms that may be understandable and accessible to the
operators potentially involved in combating desertification;
2. support the institutions involved in the designing of NAP, SRAP and RAP ;
3. effectively connect the operators involved in combating;
4. disseminate, make accessible and circulate information and products aimed at and
useful to combating desertification;
5. assist the various interested and involved partners with the NAP, SRAP and RAP
processing, setting up and follow up operations;
6. assist the planning and decision-making as regards combating desertification.

The area of application of the Project includes the Sahel, Northern and Eastern Africa.
Application is planned at the national scale, as regards the whole region, plus a number of
national systems as regards Mali, Senegal and Benin. SID Senegal is the only operational
one.
The OSS-SID Internet site is available for the visualisation of charts, it allows the
superposition of different thematic information, however neither allows to carry out analyses
nor to download information plans under the GIS format.

23

6 Data Dissemination Systems

The Data Dissemination Systems represent a source of data or processed data that are
currently exploited by the Early Warning Systems and the Information Dissemination
Systems. In fact, it could happen that the same product generated by a Data Dissemination
System is diffused more widely by an Information Dissemination System.

Most of these Data Dissemination Systems are databases mostly based on physical data,
focusing on satellite and morphological information on soils, integrated by baseline data as
well as by data on the population.

It must be noticed that AFRICOVER, of FAO, is under construction and even if its data are
not available yet, the methodology has been formulated and the project is already in
operation. Moreover, as regards FAO, FAOSTAT can also be useful vis-à-vis data and
statistics at the national level. Obviously the FAOSTAT information is limited to the
economic aspects of the market production as well to the social aspects vis-à-vis the
nutritional and the population data. The UN Africa Nutrition Database Initiative (ANDI) is
also to be considered when dealing with the socio-economic aspects.

Table 6
Data set name Data available Data format Data source
ESA IONIA 1 km AVHRR Global Land Data Set NOAA
Image
Fire Atlas Data ESA
ESRI DCW Geo-political
Socio-economic
ArcView
Climate ESRI
format
Soils
Vegetation cover
EUMETSAT Meteosat images EUMET
Images
Meteorological data SAT
FAOSTAT Economy and nutrition Tabular FAO
NOAA SAA Remote sensing AVHRR, Images NOAA
TOVS, SSM
NOAA/NASA Remote sensing NDVI 8 km
Images NOAA
Pathfinder NDVI 1 degree
PENN STATE Population
UNIVERSITY Base maps
ArcInfo export
DCW Data Server Hydrology ESRI
format .e00
Digital Elevation Model
Land Cover
UN ANDI Economy WB FAO
Nutrition Tabular UNICEF
WHO

24

Data set name Data available Data format Data source

UNEP/GRID Global Ecosystem
Global
Vegetation
Veg.Cultivation.Albedo UNEP
Data
Global Methane from
Livestock
Global Soil Texture Data
Soil Data Human Induced Soil
degradation Raster and
General Climatic Life Zone Vector
Climatic Long Term Climatic
Data Averages
Surface Crustal Temp
Global Elevation (5 min) USGS
Topographic Global Elevation (10 min)
Data Global Wetlands Database
Population Distribution Database
USAID CARPE Population data National
Vector and
Vegetation Maps Image Maryland
Uni.
Climate Rainfall Florida Uni.
Protected Areas WCMC WRI
Remote Sensing NDVI NOAA
Land cover GRID
USGS, EDC DAAC Land Cover Characteristics Data Base
USGS Land Use/Land Cover Scheme
Seasonal Land Cover Regions
Simple Biosphere Model Scheme
Simple Biosphere 2 Model Scheme
Vector and
Global Ecosystems USGS
Image
International Geosphere Biosphere
Program
Biosphere Atmosphere Transfer Scheme
DCW Urban
Digital Elevation Model
GTOPO30 Global 30
Arc Second Elevation DEM 1 KM Image USGS
Data Set
HYDRO 1K Streams, Drainage
Elevation basins, ancillary layers Image and
USGS
Derivative derived from digital Vector
Database elevation
25

Data
Data set name Data available Data source
format
USGS & NOAA: Satellite/Image Data AVHRR NDVI
Image NOAA
Africa Dissemination Rainfall estimates
Data Center Digital Map Data
ADDS
Administrative Rain Stations
Boundaries Reference Maps
Agro-Climatic Zones Roads Image and
Cropland Use Intensity Vegetation Vector
Digital Elevation West African CIA
Model Spatial Analysis OALS
Hydrology Project
Railroads
Tabular Data/Statistics
Agricultural Statistics
Tabular
Precipitation
Prices
USGS LANDDAAC NDVI 1 Km Image NOAA
USGS Global Land Climate Land cover
Information System Elevation Satellite imagery
Images USGS
GLIS Geology Soils
Vector NOAA
Hydrology

USGS Global Land Land Cover USGS
Cover Characteristics Global Ecosystems Image NOAA
GLCC Digital Elevation Model NASA
CEDARE GIS Soil Vector ACSAD/UN
ESCO
Hydrology Vector National
Basemaps Vector National
Population Land use Vector ESRI
Dem Vegetation Vector ESRI
Land cover Vector ESRI
WMO GCOS Climate and meteo WMO
ECMWF Meteo forecast ECMWF
WCMC Global African forests and protected areas.
Overview of Forest African Ecological Zones, forest cover and Maps
Conservation protected areas.
WCMC
Forest type
Total area of forest in each country Tabular
Forest area in each ecological zone

26

The USGS proposes a series of data sets with emphasis both on remote sensing data and on
their processing, such as land cover maps. The NOAA Satellite Active Archive (SAA)
clearly focuses on the global cover of the NOAA products; processed data are also available,
still they mostly concern the national level. The Pennsylvania State University and the ESRI
are worth mentioning as they offer access to the Digital Chart of the World.

The field of application is varied enough as it ranges from the global to both the national and
local levels. The GRID and Eros Data Center, for example, have a more extended
application, i.e. the global level, while ADDS is limited to the African area, as the acronym
clearly shows, and the CEDARE is structured on a country basis.

Graph 5 Available data
B as e m a p s C lim a t ic D a t a
W ater res o u rc e s 7% 12%
7%
Econom y
9%

V e g e t a tio n S oil D a t a
28% 15%

P o p u lation R e m o t e s e n s in g
11% 11%

7 Early Warning System and desertification

Early Warning Systems are largely based on bio-physical data and do normally perform
also environmental evaluation finalised to the assessment of food security. The list includes
vegetation evaluation/monitoring, by means of remote sensing, agro-ecological or agro-
climatic zoning, land use and land cover classifications etc..

The existing Early Warning Systems utilise environmental and socio-economic data and
indicators, which could be directly employed to assess land degradation or to monitor
desertification. The EWSs are deeply focused on food security and they are approaching
other fields of application just occasionally, more for institutional reasons than for technical
causes, in order to avoid any possible conflict between traditional and new stakeholders.

Therefore, only an increasing pressure from users could facilitate an extension of the field of
activity from food security to natural resources management.

Desertification is an element which affects the eco-systems, that are generally monitored by
the Early Warning Systems, in fact it is together a cause and an effect. Soil degradation is
the result of natural processes that are either induced or catalysed by man. It produces the
deterioration of the vegetation cover, soil and water resources. Through a series of physical,

28

chemical and hydrological processes, this deterioration causes the destruction of both the
biological potential of lands and their capability to sustain the populations involved.

Land degradation being together a cause and an effect of food scarcity, it is generally
indirectly monitored by the EWS. Due to its intrinsic correlation with human and socio-
economic factors, desertification could be measured also by means of the same
methodologies utilised by the food security.

Nevertheless, some important distinctions must be made. The temporal scale of a food
security EWS is basically conditioned by the rapidity of evolution of the processes under
examination, however, the desertification processes take place in the medium- and long-
term period. The effects of the climatic changes and of land degradation are too slow to be
included into a risk analysis such as the crop or agro-pastoral analysis. As a consequence,
the time scale of an EWS for desertification should be extended over a period of several
years in order to see remarkable changes in evidence.

Two items seem to be faced with the most serious impact:
§ the micro-scale level analysis of the effects of the populations/environment dynamics
in i) the areas where desertification processes are particularly rapid, ii) those with
population migrations and iii) those where modifications of the productive systems
occur at the same time as remarkable climatic changes;
§ at the regional and global level, the assessment of the status of desertification, enabling
the analysis of the changes occurred in the last decades, for i) the quantitative
assessment of the desertification extension and ii) for the identification of the degree of
vulnerability.

8 Conclusions

As of today, the Early Warning Systems are deeply evolving, due to the changes occurred in
the technological environment in which they operate. However, this process should take into
account the present or potential end users, who are facing as well a new information
technology and communication world.

In the near future, some key questions are arising on how to set up a real demand-driven
EWS, rather than a system developed just under the pressure of a technological push.

Need to develop a common language. The integration between a risk analysis and a
vulnerability analysis, as the structural frame of reference, has become a generally shared
approach. The different meanings of particular terms, i.e. vulnerability and risk, in different
systems is still misleading, making interaction difficult and causing isolation.

29

Facilitated access to and transparency of data. As of today, access to baseline data, in
particular, is really neither free nor facilitated, due both to the difficulties in making the data
banks’network operational and to the idea that data collection would be the final objective.

Accelerated interaction towards the real partnership. A complex system requires – especially
at this stage –the real willingness to co-operate with a partnership attitude, vis-à-vis those
institutions that might contribute to its development and the donors/agencies who are asked
to establish a political and institutional ‘enabling environment’.

Production of a focused information for decision-making. At present, the conceptual
capability of interpreting information is still behind the information production potential,
and the risk arises that an unfocused information will be generated. This would charge the
user with the task of selecting the information, rather than commanding it.

Users are required to identify the information they need. Users are not a homogeneous
category, as regards both their technical skills and their information demand. This is
certainly a further difficulty facing the EWSs that must decide, without any active interface,
the type of information that is to be provided.

Adequate development of national/sub-national nodes. All the systems under consideration
are operating at regional or sub-regional level, even if they produce information at the
national or local level. How can any national and local EWS be functionally and
institutionally developed, so as to be introduced into the existing network of systems like
those under consideration?

Acceleration of the passage from food security to security. All the early warning systems
under consideration are expanding to new fields of concern, such as economic planning and
management of natural resources. Under this aspect, the vulnerability analyses are moving
towards an improved interaction between environmental and socio-economic classifications.
In this regard, the attention is being drawn on the concept of ‘security’ which is based on an
,
organic complex of data and different only as regards the analysis’path.

Technological development should not be considered as a priority. The information
technology is sharply and quickly developing. New generations of satellites are rapidly
becoming operational. Therefore, the EWSs are endowed with theoretically more and more
powerful tools. In this framework, it is of capital importance that - with respect to these new
tools - priority be given to the development of those applications that would be really suited
to the end users.

30

Early Warning Systems

Recommandé

Recommandé

Contenu connexe

Tendances

Tendances (14)

En vedette

En vedette (20)

Similaire à Early Warning Systems

Similaire à Early Warning Systems (20)

Plus de kaakaawaah

Plus de kaakaawaah (20)

Dernier

Dernier (20)

Early Warning Systems