STRENGTHENING DROUGHT EARLY WARNING AT THE COMMUNITY AND DISTRICT LEVELS: Analysis of Traditional community warning systems in Wajir & Turkana counties

A report to Oxfam GB, Kenya

1

Authors’ affiliations and mail addresses Sammy Mutua, DRR Consultant TASK Consulting Services [email protected] P.O. Box 27368-00100 NAIROBI

© Oxfam GB 2011

Oxfam GB, PO Box 40680, 00100 Nairobi, Kenya Shelter Afrique House, Mamlaka Road Tel: +254 (0) 20 2820000 Fax: + 254 (0) 2820104/5 Email: [email protected]

Disclaimer: This report has been produced with financial support of the European Commission (ECHO). The views expressed in the report should not be interpreted in any way to reflect the opinion of the European Union.

2

Table of Contents 1.0.

INTRODUCTION

1.1. Background 1.2. Purpose of the study 1.3. Objectives of the study 1.4. Methodology 2. 0. CONCEPT AND THEORETCAL FRAMEWORK OF EW 2.1. Drought Warning Systems 2.2. Importance & uses of EWS 2.3. Examples of Early Warning Systems 3.0. DROUGHT EARLY WARNING SYSTEMS (DEWS) 3.1. Components of DEWS 4.0.TRADITIONAL KNOWLEDGE IN DROUGHT MANAGEMENT 5.0. THE EWS SYSTEMS IN KENYA 5.1. Arid Lands Resource Management Programme 5.2. Project core functioning 5.3. Drought Management Initiatives 5.4. The role of ALRMP in Early Warning

6.0. EXISTING TRADIONAL EWS 6.1.Wajir communities 6.2. Seasonal weather forecast 6.3.The Turkana County 6.4. Adaptive strategies of Turkana county 6.5. Traditional Early Warning Indicators 6.6. Early warning indicators from 5 communities in Turkana 7.0. TRADITIONAL EWS TO INFORM DISTRICT LEVEL ANALYSIS 7.1. Channels for EWS, TEW & scientific EWS 8.0. RADIO MESSAGES FOR COMMUNITY LEVEL EWS 8.1. Area specific Radio messages 8.2. Proposed Hybrid Early Warning Dissemination process

9.0. CONCLUSION

10. REFERENCES

6 6 6 6 6 8 9 9 9 11 13 13 16 17 17 18 21 22 22 23 25 26 29 31 32 36 38 40

3

GLOSSARY TEWS HEWS DEWS ALRMP UNISDR DCF EWB OXFAM KMD EWS FGD DEWS DEW GOK DOL EW TK DSG NGO DMI DWEB CMDRR PVCA DRR ICPAC DMO DPA LMA NDCF FM KFSSG MoSDoNK NRM

Traditional Early Warning systems Hybrid Early Warning System Draught Early Warning System Arid Lands Resource Management Programme United Nations International Strategy for Disaster Reduction Drought Contingency Funds Early Warning Bulletin Oxfam Great Britain Kenya Metrological Department Early Warning Systems Focused group discussions Drought Early Warning Systems Drought Early Warning Government of Kenya Diocese of Lodwar Early Warning Traditional Knowledge District steering Group Non Governmental Organization Drought Management Initiative Drought Early warning Bulletin Community Managed Disaster Risk Reduction Participatory Vulnerability Capacity Assessment Disaster Risk Reduction IGAD Climate Application Prediction Center Drought Management Officer District Pastoral Association Local Marketing Associations National Drought Contingency Fund Frequency Modulator Kenya Food Security Steering Group Ministry of State for Development of Northern Kenya National Resource Management

4

ACKNOWLEDGEMENTS The study on Traditional Early Warning Systems was launched by Oxfam GB and was conducted in April and May, 2011 covering Wajir and Turkana Counties. The consultant wish to thank all those who contributed during the field visits and final production of this report. This includes those at both national and district level that were willing to put aside their busy schedules and share ideas and sometimes answer what may have seemed too inquisitive questions. Special thanks go to Oxfam GB DRR Manager and his team both in Wajir and Turkana for their enthusiasm, support, accompaniment and understanding during the assignment. Your input and thoughts were extremely valuable. Finally and very important, special gratitude go to the pastoral communities in Wajir and Turkana counties who in the midst of a drought crisis managed to create time and share their wealth of knowledge and experience on their own traditional drought Early Warning systems that have kept them going for decades .

5

SUMMARY Oxfam has supported local communities in two counties of Wajir and Turkana to identify their drought risk profiles and develop disaster management plans. At the same time, there is a corresponding effort to strengthen drought early warning systems (DEWS) among pastoral communities. The main thrust in this is the presence of clear drought monitoring indicators that can be monitored over time and appropriate to trigger signals relayed to the relevant institutions for appropriate actions. The two counties have an exiting drought early warning system managed by ALRMP. The ALRMP prepares a Monthly drought early warming bulletin (EWB) that is disseminated to stakeholders. Unfortunately getting this information to the communities has been a challenge, besides the packaging of the EWB is not friendlily to the pastoralists because of the high illiteracy levels among other issues. The aforementioned challenges have contributed to low uptake of the EW information leading to insufficient drought preparedness among pastoralists. In seeking ways to strengthen the existing EWS, this study interrogated the traditional early warning systems (TEWS) of both the Somali and Turkana communities in Wajir and Turkana counties. The study also analysed how the TEWS of these communities can enrich the scientific EWS generated by ALRMP and Kenya Meteorological Department (KMD) and by so doing increase the uptake of the EW information by pastoralists. After extensive consultation with various stakeholders and communities from the two areas, a hybrid early warning system (HBEWS) for drought management was proposed for consideration. However, it is important to note at this early stage that the proposed HBEWS will NOT replace the existing scientific EWSs, but rather a system that is informed by both the scientific EW information from both KMD & ALRMP and traditional early warning information. The system should be managed by ALRMP and to increase the uptake of EW information, the study proposes use of local FM radio stations. It is expected that the integration of the traditional and scientific EW systems will enhance the trust of communities on early warning information and increase its uptake.

6

1.0. INRODUCTION 1.1. Background of the study OXFAM has been working in Kenya since 1963. Oxfam is currently implementing a drought preparedness project in Wajir and turkana counties. The project focuses on development of community disaster management plans as well as strengthening of drought early warning at the community and Districts levels. Early Warning (EW) is the provision of timely and effective information through identified structures that allows individuals exposed to hazards to take action to avoid or reduce their risk and prepare for effective response. Kenya has a national drought early warning system that has been applied at the district level. Oxfam have been supporting communities in both Wajir and Turkana to identify their drought risk profiles and develop disaster management plans at the community level. Nevertheless, some gaps have been identified in the way the current EWS is applied at the district and community levels. For example, the current system mainly focuses on scientific information to monitor drought despite the presence of a wealth of knowledge among the pastoral communities in dealing with climate variability over the years. Perhaps the current EWS could become more reliable if fused with existing traditional EWS that have been used by local communities. As a starting point, there was need to carry out a participatory scoping study of these traditional EWS in order to tease out useful indicators that can reinforce the existing conventional drought indicators. 1.2. Purpose of the Study The overall aim of this assignment was to interrogate the existing traditional drought EWS utilized by communities in both Wajir and Turkana districts to ascertain if these systems could be linked to the conventional scientific drought EWS (currently being utilized by ALRMP) in order to strengthen the precision of EW and consequently enhance drought preparedness efforts among the targeted communities in the two counties. 1.3. Objectives of the study To achieve the above purpose, the following is a summary interpretation of the objectives of the broad assignment: i. To carry out a participatory scoping analysis of existing traditional EWS ii. To enhance the linkage between traditional EWS and district level drought analysis iii. To develop simple , culturally acceptable and timely radio EW messages

7

1.4.

Methodology

The strategy for collecting information on traditional EW knowledge in this study comprised of a mixture of various research techniques which included the use of interactive interviews, Focused Group Discussions (FGDs), guided semistructured interviews, documentation from oral history, reports, research publications, and past research studies on indigenous knowledge on drought early warning. The consultant first held meetings with key people at the district level (involved in drought management and users of the ALRMP DEWB) before proceeding on exhaustive field visits, in Wajir and Turkana Districts. At the district level, meetings were held with key stakeholders that included officials from key Government line ministries and NGOs. Their views were sought and analyzed in order to get in-depth understanding of the scientific DEWS currently being used in the two counties. Their views regarding the traditional DEWS were also sought before proceeding for further consultations with communities in the field. Four (4) community sites (two in each area) were selected and visited on the basis of areas targeted in the Oxfam ongoing DRR interventions. More specifically, the consultant visited the following sites: Biyamadhow & Sarif and Wajir Bor in Wajir County and Lokore and Lokangae in Turkana County. In each of the 4 project areas visited, the consultant was accompanied by either Oxfam field staff or a staff from the respective Oxfam partner on the ground. Again this team was expanded to include local leaders and community members from each specific community visited. Two communities where visited in each area and extensive FGDs conducted with each group. Each meeting ensured adequate inclusion of women and youth was catered for, their opinions and input consulted and captured during the FGD process.

8

2.0. CONCEPT AND THEORETCAL FRAMEWORK OF EARLY WARNING 2.1. Early Warning System (EWS) There is no universally agreed definition of a EWS. However, for the purpose of this study, a EWS shall be referred to as a system of data collection that facilitates detection and monitoring of impending hazards so that actions can be taken to reduce the negative effects of these hazards. It is a system of data collection to monitor people‟s access to resources and protection in order to provide timely notices when a crisis threatens and thus elicit appropriate responses1. According to the United Nations, EWS is defined as the timely provision of effective information, through identified institutions, that forewarn individuals exposed to a hazard to take appropriate action to avoid or reduce their risk and prepare for effective response2. However, the EW information should be sensitive and specific to the underlying phenomenon for it to be meaningful. It can thus be deduce from the above definitions, that EWSs are the procedures and actions through which information is produced in advance about the occurrence of potential hazards. Early warnings are therefore, public services that responsible government authorities must provide to communities to minimize future potential risks. Important to note is that, EWS are based on incomplete and probabilistic information which means, judgment and community involvement is extremely important to enhance ownership and trust of the EWs information. Several activities are involved in an early warning process. These includes; data collection, information development, dissemination methods and action triggering mechanisms. Thus the EWS process should be understood in the context of an integrated and holistic risk management framework3 . Moreover, cross cutting themes such as good governance, institutional arrangements, multi-hazard approaches to EW, involvement of local communities, gender and cultural diversity are also fundamental in the development, use and determination of the most appropriate EWS.4 Besides, for an EWS to be effective, it must be able to trigger timely response, - thus intervening before the crisis point has been reached. EWS must be able to protect livelihoods before lives are threatened, and must be seen to be geared to protect future capacity to subsist as well as ensuring current social well-being of the 1 2

3 4

Stephenson, 1994; Monnik, 2000; Buchanan-Smith, 2000. UNISDR, 2003.

Hab, Asch et al, 2009. UN/ISDR and EWCIII, 2006.

9

communities at risk. EW should therefore be able to detect localized pockets of severe stress and trigger signals for early action. In a EWS, four phases can be identified and followed. These are presented figuratively as below: 1) Measurement of Precursors

2) Forecast indicates catastrophic event? If yes go ahead to warning, if no, back to measurement of precursors

3) Warning

4) Onset of anticipated response

2.2. Importance and uses of EWS The aim of early warning is to empower individuals and communities, predisposed to hazards, to act in ample time and in an apt manner so as to reduce the prospect of personal injury, deaths and damage to property, or surrounding environments. Crucially, since EW is a major element of DRR, it is important in development endeavors. Besides preventing loss of lives and reducing negative impacts of disasters, EW help people to deal with potential disasters and aid in the process of recovery and sustainable development5. EW should seek to provide decision makers at all levels with information on the onset, continuation and end of a disaster status. Indeed, EW provides information on vulnerability factors and patterns and hazard forecasting. Better identification of risk and occurrence of hazard aided by strong monitoring of vulnerability levels, can be achieved through EWS. In the Hyogo Framework for action 2005-2015, which is the roadmap negotiated by governments at the World Conference on Disaster Reduction in Kobe, Japan in 2005, DRR priority number two is the EWS. In the framework, there are clearly 5

Environment and Societal Impacts Group, 2004; UN/ISDR and EWCIII, 2006.

10

spelt out activities under this priority, which are: Risk assessments, early warning, capacity enhancement and focusing on regional and emerging risks6. EWS products can include: - Short to medium term weather forecasts - Prediction of the onset and cessation of a rainy season - Prediction of wet and dry spells - Climatic forecasts and soil moisture monitoring - Outlook for onset of drought. During the field visits, communities expressed two areas of concern. One, although they follow the EWS information relayed by Kenya Metrological Department (KMD), they end up losing their livestock before any assistance from the government arrives. This undermines the effectiveness of the existing EW which depends not only on technical capabilities but also on the preparedness of decision makers and their immediate response. Second most warning messages sometimes difficult to understand and this underlines the need for simple warning messages that are understandable and are accompanied with clear reaction instructions within guaranteed time frame7. More often than not, the concept of managing disasters is reduced to hazard assessment and emergency response. Great value is placed to the scientific understanding of hazards and protective structures, while analysis of socioeconomic impacts and risk assessment plays second fiddle. However, this is another domain where EW should play a role to better understand the communities under threat and how they are or can be prepared to respond 8 to a potential hazard before a disaster strikes. Ideally, an EWS should not be a process of data collection and analysis that is considered the end result. Rather, it ought to be part of a bigger system that is aimed at mitigating and responding to a crisis. Notwithstanding, EW should also be sustainable through innovative and adaptive technologies, scientific knowledge and by review of their aims and performances9. This provides an appropriate opportunity to review the value of bringing on board the traditional early warning knowledge of communities to be part of the bigger EWS system referred above.

UN/ISDR, 2005; Chung, 2005; Wong, s.a. Samarajiva and Waidyanatha, 2007; Hab et al, 2009. 8 Hab et al, 2009. 9 Monnik, 2000; Benfield Hazard Research Centre, 2006. 6 7

11

There is no doubt, EWS eventually pay off. A prerequisite for an effective EWS is the recognition of its benefits by the general public, policy makers and the private sector since EWS have in fact turned out to be practical tools for implementing timely and appropriate responses to hazards. 2.3. Example of Early Warning System An example of a EWS initiative is the Famine Early Warning Systems Network (FEWSNET). This USAID-funded initiative covers 17 countries in sub-Saharan Africa and Afghanistan in Asia. The network offers a range of information products, tools and services aimed at strengthening the abilities of the countries and regional organizations to manage threats of food security through the provision of timely and analytical early warning and vulnerability. It also provides decision makers with up to date information to avert or mitigate the impacts of food security shocks. The challenge has been to ensure EW result in prompt responses by governments and potentially the international community. It also requires that the information is effectively disseminated down to the end user especially rural communities via a media system that is accessible and in a language which is locally understood.

12

3.0. DROUGHT EARLY WARNING SYSTEMS (DEWS) Drought monitoring and EW are major components of drought risk management whose goal is to increase the society‟s coping capacity to enhance greater resilience and reduce the need for disaster assistance . DEWS are systems of data collection for monitoring drought indicators and are supposed to provide timely notices for appropriate community drought responses10. A DEWS is designed to identify climate and water supply trends and thus detect the emergence or probability of occurrence, onset, development, persistence, alleviation, severity and the end of a drought. This information can reduce the impacts if delivered to concerned stakeholders in timely and appropriate formats. However, mitigation measures and preparedness plans need to be put in place in order to minimize the prior impacts11 of the hazards. An example of DEWS is the USAID supported Famine early warning system, which encompasses monitoring the situation in areas known to be particularly vulnerable to the effects of drought, crop failures or changes in economic conditions, to enable remedial measures to be initiated before hardship becomes acute. It is reiterated that drought is slow onset in nature and EWS must have the capacity to detect the early emergence of rainfall deficiencies, which is normally the best indicator of incipient drought periods12. It is necessary for effective DEWS to integrate precipitation and other climatic parameters with water information such as stream flow and soil moisture into a comprehensive assessment of current and future drought and water supply conditions. With minimum orientation, apparently this sort of information can be availed easily by local communities themselves. Thus, A DEW needs simple but dynamic methods and empirical evidence. Since drought creeps in slowly, it is difficult to quantify and therefore needs multidisciplinary variables/indices and the monitoring of the whole range of sectors: agriculture, water, health, energy among others.13 . 3. 1. Components of DEWS Early warning in slow onset hazard such as drought can be a resilience-building tool. In many countries, particularly Africa, DEWS are also coupled with those developed for famine and food shortages and they generally become indicators of stress on lives and livelihoods. UN/ISDR, 2003. Kerang, 2000; Action aid, 2006. 12 UN/ISDR, 2002. 13 WMO, 2006; De Pauw, 2000. 10 11

13

DEWS must not only encompass mechanisms and procedures for the collection and analysis of information in a timely manner but also for dissemination of that information through locally appropriate channels to end users. End users should be made aware of the essentiality of this information and continuous information flow should be encouraged once a drought is foreseen14. Further, DEWS are an indispensable component of drought preparedness plans and policies and practitioners including disaster managers must work together to design products that better communicate information to decision makers in order to reduce the risks associated with drought. As always, input from end users should be solicited to better understand their information needs. Below is an illustration of a typical scientific DEW chain:

Observing networks Synoptic stations Hydro-meteorological stations Remote sensing data Aerologic data

Forecasting models

Warnings

Source: Eljohra, 2006;

Due to drought‟s slow onset characteristics, monitoring and EWs provide the foundation for an effective and responsive drought mitigation plan. Typical drought plans normally contain three (3) basic components: a) Monitoring and EW, b) Risk assessment and c) Mitigation and response. The drought plan however, must rely on accurate and timely assessments to trigger contingency and emergency response programmes15. Thus , DEWS do not work alone and it requires the presence of pro-active community drought contingency plans and a reliable source of „adequate‟ funding (drought contingency funds) in order to ensure timely execution of both emergency response interventions to ensure realistic impact on the lives of the pastoralists is achieved.

14 15

UN/ISDR, 2003: 5:para 3). Wilhite and Svoboda, 2000).

14

Inadequate funding for contingency plans is a reality with serious loses to pastoralists. Oxfam has done a separate study on drought contingency funding. To explain the gravity of the existing funding gaps is a case encountered in Turkana County during one of the field visits. During the period of this study, the GOK was running a drought contingency programme for livestock off take. In Lorau village, Nanam Division of Turkana West, the community members here complained bitterly that due to inadequate funding of the programme , only 100 shoats were slaughtered out of the thousands of shoats the pastoralists had gathered ready for slaughter in their location („too little too late‟).16 Some DEWS focus on the hazard of the impending disaster and not on the vulnerability of systems and communities. Information on vulnerability is required to provide a focus or drought interventions. A vulnerability profile can complement the EWS and aids decision makers on how and where to respond for maximum effectiveness, for example, recent rainfall levels, prices, nutritional status, environmental status, soil fertility and household status.17 4.0. TRADITIONAL KNOWLEDGE IN DROUGHT MANAGEMENT Traditional early warning systems represent the whole body of knowledge developed early in the 1900s among pastoral communities to anticipate the coming of rains and thus enable them to mitigate the effect of droughts on their livestock - the backbone of their livelihoods18. There is need to appreciate traditional early warning systems and drought coping strategies which have helped pastoral communities in Turkana and Wajir to survive for generations. The most effective way to collect and transmit such local information for use by early warning systems remains to be worked out and should be made a priority for continuous review and research and discussion19 especially with variability of weather caused by climate change. The role of traditional knowledge in participatory approach has 2 main values. First, using TK employs information which is already possessed by the community and which is also valuable to DRR. In both Turkana and Wajir counties, the local communities have a history of experience with drought disasters. These communities have accumulated a wealth of experience and information regarding prediction, reaction and recovery from drought and this The disappointed pastoralists complained bitterly about the big publicity of the GOK livestock off take programme without matching funding to meet the high livestock numbers that are too week to survive the drought. 17 Monnik, 2000. 18 C. Platt: 2001. 19 Rebecca H. A., u.d: 8. 16

15

provides a good opportunity for potential partnership with scientific weather prediction initiatives. The recognition and use of TK can provide improved self confidence for the community and allow it to deal with disasters on its own. Recognizing and sharing TK will confirm to communities that their knowledge is valuable and give its members authority over the process of risk reduction. This in turn will provide the enhanced security needed to immediately respond to potential threats from disasters since local community members are the first responders. There is a strong need to strengthen the resilience and self confidence of local communities to cope with drought through recognition and propagation of traditional early warning practices and values as part of DRR activities. Further, there is evidence of disappearance of some of the traditional strategies employed by pastoralists to mitigate the effects of drought that have evolved over the years under very different conditions. This is cause for concern because the number of emergency responses/options available to communities has been reduced. Conditions that exist today, particularly access to relief food or other resources supplied by the government or external agencies, may change in the future. Likewise, climatic or other conditions may improve or deteriorate. Given these uncertainties, it is important that these communities have as many choices as possible to employ in their overall survival and livelihood strategies hence the need for preservation of these practices in some form for future generations20 . 21In

Wajir South for example, elders pointed out civilization, religion and education as some of the key developments that have contributed to limited adherence to the traditional drought coping mechanisms including the traditional early warnings.

The processes of urbanization, development, formal education and changing religious ideologies have all contributed to the decreasing practice of traditional early warning systems and coping strategies. Because this knowledge is in many cases preserved as an oral tradition, the increasing absenteeism of youth from pastoral communities and the decreasing practice of traditional knowledge pose a great threat to the continued existence of the traditional of experience (Dr Pratt, 2002). 21

16

5.0. DROUGHT EARLY WARNING SYSTEMS IN KENYA 5.1. The Arid Lands Resource Management Project (ALRMP) – Scientific EWS The Arid Lands Resource Management Project (ALRMP) which is under the Ministry of State for Development of Northern Kenya (MOSDONK) is the official government agency responsible for Drought Early Warning in the country. The programme currently is covering 28 districts in Kenya. The ALRMP is a community-based drought management project of the Government of Kenya (GOK) that utilizes a credit facility from the World Bank. In 1996, the ALRMP was initiated after the World Bank-funded Emergency Drought Recovery Project (EDRP), which operated from 1992 to 1995, came to an end. While the EDRP was a quick-fix project to mitigate the effects of severe drought, ALRMP was a longer-term development-oriented project focusing on institutionalization and mainstreaming of drought management activities in the Kenya Government system. The first phase ended on June 30, 2003. The achievements of ALRMP 1 were considerable but it was not possible to move towards full institutionalization of processes due to the nearly persistent drought emergency situation during the implementation period. It is because of this, among other reasons, that the second phase of the ALRMP was proposed. ALRMP II was launched in September 2003, with initial coverage of 22 arid and semi-arid districts, and was to be implemented over six years (2003-2009). The Project suffered the effects of a severe drought in 2005, which saw the Government spent enormous resources on emergency measures such as water trucking, livestock vaccination and treatment, drought-related human health and nutrition interventions, prevailing disease outbreaks and emergency livestock off-take. The World Bank supervision mission witnessed the effects of the drought during the 2006 ALRMP II midterm review. The Bank then gave 'retroactive funding' to the GoK as a form of compensation for the funds some GoK ministries had relocated from development activities for use during drought emergency; and additional funds to boost ALRMP II activities, including the Drought Contingency Fund (DCF) kitty. The funding enabled expansion of ALRMP II to an additional six semi-arid districts, raising coverage from 22 to the current 28 arid and semi-arid districts. The project was also extended by one year (2003- 2010) to provide adequate time for utilization of the additional fund

17

5.1.1 Project core functions The core functions of the ALRMP have being implemented through the following components: Drought Management Natural Resource Management Community Driven Development Support To Local Development

5.1.2 Drought Management Initiative Development partners have been supporting ALRMP activities. One such support is from the EC through the Drought Management Initiative (DMI) Programme initiated in 2007. The DMI was a four- year project implemented within the ALRMP II framework. DMI‟s objective is to contribute to the improved effectiveness and efficiency of the drought management system in Kenya by strengthening the capacity to intervene in an appropriate, effective and timely fashion during the drought cycle. Some key DMI activities that support ALRMP drought management activities include: -

-

-

-

-

Provision of additional financial resources for the existing DCF. Improvement and harmonization of the regulations, administrative and financial procedures governing the DCF and its transformation into the NDCF. Establishment of close coordination mechanism between the Drought Contingency Fund DCF/NDCF and other drought management structures and institutions and provision of support to the Early Warning System. Building on the ALRMP process of legalizing the role and the capacity building of the District Steering Groups (DSGs) as key components of drought coordination and early warning at the district level. Fostering the formulation and adoption of policy documents moving towards the formalization of drought management structures. Setting up of a flexible disbursement grant system with locally based implementing agencies receiving funds from ECHO to implement recovery programmes that Link Relief Rehabilitation and Development (LRRD) activities in order to ensure a smooth transition from the emergency phase to the recovery phase including disaster Support the drought preparedness mechanisms of the ALRMPs.

18

5.1.3 The Role of ALRMP in Early Warning The Arid lands Resource management programme (ARLMP) is the agency mandated by the GoK to collect analyze and report on EW information. The organization has deployed data collection monitors based in the community every month, collect household data and sent it to the coordinating office based at the district level. The data is analyzed and drought monitoring bulletins produced. All NGOs and government agencies at the District and National levels rely on this bulletin known as Drought Early Warning Bulletin (DEWB) which is produced by the ALRMP for early warning information. Wajir and Turkana counties which are the focus of this study produce their own monthly DEWB which are shared with relevant stakeholders in the districts. The ARLMP also co-ordinates and plans interventions through the District Steering Groups (DSGs). 5.2. EXISTING TRADITIONAL EW SYSTEMS IN WAJIR AND TURKANA Traditionally, most communities in Kenya have indigenous systems and mechanisms of dealing and coping with drought. These systems include traditional indigenous weather and climate prediction practices that are based on indicators established over generations through keen observation of animals, birds, insects, the solar system, winds, clouds, and the feelings of the human body. The communities recognized unique situations associated with the behavior of the above living organisms, and the locations and patterns of cloud, winds, the moon and stars. The predictions based on these indicators and human feelings supported the early warnings issued by the elders to enable the community cope with the anticipated natural hazard. 22Traditional

-

-

22

early warning systems for pastoralists are based on three precepts:

Detailed knowledge of exact arrival of the major rains season - understanding the forecasting and the probability of future rain by based on variance in wind, humidity and temperature from expected conditions. Pastoralists interpretation of both animals and plants behavior, which serve as valuable indicators for subtle fluctuations in temperature and humidity. Appropriately interpreted, local pastoralists can forecast major rains four weeks before their arrival. Similarly, the absence of these natural indicators suggests the absence of rain, i.e. drought.

Dr. Platt: 2002.

19

-

Observed historical trends allow for reasonable predications of future weather patterns. However, the increasing severity and frequency of drought over the last few decades have rendered this latter form of forecasting less reliable than it was in the past.

To put into practice some of the traditional drought mitigation practices based on the understanding of climatic changes as described above, both the Somali and Turkana communities have relied on traditional EW indicators to predict future weather patterns for their survival. The consultant established that the most trusted traditional EW indicators from both communities (Turkana and Somali) are connected with domestic animals 23. Moreover, the communities confirmed that, livestock related indicators are the most dependable and as such, most likely to set in motion proactive drought mitigation actions by the two communities. The following is a summary of traditional EW indicators derived from focused group discussions with community members from the two areas: 6.1. Wajir Communities Two sites were visited, Biyamadhow (in Wajir South) and Wajir Bor (Wajir West District) and extensive FGDs held with community representatives. The FGD meetings brought together community members of different age sets of; elders, women, and youth. The communities visited in Wajir explained that their Somali traditional culture is quite similar to the Borana communities that inhabit large areas of Moyale and Isiolo and this means most traditional EW indicators cut across large geographical areas. The following are examples of traditional signs and indicators of predicting coming seasons among the Somali communities in Wajir: Early Warning (Animal behaviors) -

-

Sounds and behavior of cattle – If the cattle refused going back to the Kraal and run away while facing different directions it is a sign of bad season while if they ran towards the Kraal/home around 3 o‟clock in the afternoon -it shows a good season is coming- rains. Grunting of the camel bulls (making mating calls) symbolizes a good rain season. The jovial mood of camel also is an indication that the coming season is good as well.

Platt (2002: 26) this is collaborated by Platt in his argument that , the close proximity and interaction of human with livestock explains why most of the indicators relating to drought come from the observed behavior of domesticated animals, particularly camels and cattle’. 23

20

-

-

-

-

-

When cattle make some specific throat vocal sounds when sleeping at night is an indicator that the coming season is going to be good. However, there is some throat vocal sounds if they make will indicate a bad season as well. When drought is looming in the coming season, cattle take excessive water and refuse/ or are reluctant to move away from the watering point. The communities in Wajir strongly believe that this behavior is an indicator of a dry season in the offing. The opposite is held true, when cattle take water and run around jovially as if in a celebration mood, it predicts good season (rains) is coming. It is believed that if cows give birth to female calves during drought period, it symbolizes that the community should prepare for a bad season (drought). The cattle bulls become „arrogantly playful‟ when the coming season is going to be good. Sometimes breaking watering troughs with their horns as if to declare that „very soon the cattle will not queue for water‟. Sometimes female camels urinate while sitting as if to express the hopelessness for the future. Sudden increase in libido, mating and general excitement of livestock symbolizes the rain season is approaching and prod communities to make upfront decisions. If camels run away, are browsing haphazardly and appear to have lost their hearing senses, it depicts a bad season. When camels brush their legs together - this symbolizes a good indication of a bumper season Firan fir – when the Mole throws up soil making the sand loose for aeration to take place, it depicts a very hot season.

Early Warning – indicators related with behavior of Insects - When black ants make and follow only 1 route from their hole is an indicator of a bad season while if they made more than 4 routes from their holes it shows a good season. Early Warning Indicators related with behavior of birds - Irr Kujira – This is a type of bird that produces sharp sounds during the night. The sound that they produce at different stages depicts whether the season will be bad or good. If this bird (Irr Kujira) makes the noise between mid night and early in the morning, this is a sign of a good rain season. Early Warning (Astronomical) -

“Jibin taladha dace tagoog geel oo biya geleen” – If the shooting star shoots on Tuesday, it is interpreted that camels will soon get stuck in mud and this symbolizes heavy rains. Enough rainfall.

21

Early Warning (Meteorological indicators) - “Deyr bula kalaheys, rati ii rii ulahaw” – this is a Somali saying which means, if the short rains (Deyr) start early in the mornings, get prepared with your camel and shoats. This means the seasonal rains are going to be good for browsers (camel and goats) but there will be no grass for cattle-(short rains normally kicks in between October - December). - “Guu bula kalaheys, sac iyo salah ulahaw” – this is interpreted that if the long rains(Guu) which sets in between April & June starts early in the morning, the community should prepare their cattle and land. Symbolizing good seasonal harvest and plenty of pasture for livestock during that season. - Failure of Lih Kora – These are out of season showers, usually expected in the month of February when the Jilaal is at 60 days i.e between Januarys and February. These showers are perceived to come in - in good time to clear dust off the environment. These showers are normally expected on the 56th Day i.e. towards the end of February. - The Failure of Todob – These are showers expected when the Jilaal is at its 70th day. These are normal in the second week of March. - Failure of Habar Ari – This is expected to be the middle of the rainy season and normally this rain is expected on the 90th day and is believed that these are the rains (showers ) that normally chase away the Jilaal (the dry season) . - Failure of the Dirir – The Dirir is the peak of the April rains and is normally expected on the 105th day which is on 15th of April 6.2. Seasonal weather forecast calendar for Wajir County The following is a seasonal weather forecast chart generated by Somali youth during a FGD meeting in Wajir and validated by Somali elders (with few amendments) from Wajir Bor- Wajir West District Season Jilaal Jan to March

Guu April to June

Weather Characteristic Indicators in Wajir High temperatures Weathering of trees and shading of leaves Drying out of water pans & Environment becoming dusty Livestock production reduces Increased water consumption Clear clouds Cracks in the grounds High temperatures at the beginning of April Nimbus clouds cover the sky & Very cold at nights Trees produce buds High dusty winds – Toloolia Anxiety of animals (animals become „jovial „) – Beeyriiq Reports of straying of camels

22

Hagaay July to August

Deyr October to Dec

Mating of Shoats – Qalaat Sahaan- Herders move out in search of livestock pasture Break out of common cold Too windy through out the season – Shurey There is rain Drizzles Livestock skin change – Cathaar Reduced animal weight Out break of diseases – Shimbir, hergeb Strong wind that dries the grass pasture Reduced pasture Malnourished children Sheep experience running noses and mucus High temperatures – Kaliil Unpredictable clouds There is anxiety among community members as uncertainty creeps in Reports of animals getting lost strong will to survive among community members - not giving up Trees producing buds and wild fruits – Mareer - Howlan NGOs start collecting Data from the community

It was clear from the FGDs that seasons represent one of the most important pastoralists‟ traditional early warning indicators. Knowledge of the seasons, when and when not to expect rains, helps to reach conclusions on the probability of drought and the subsequent course of action on the part of pastoralists. According to the Turkana community members interviewed, the reliability and accuracy of a particular indicator is measured by its confidence weight value. However, this does not mean if an indicator has a higher confidence value based on its strength as a drought prediction is the best indicator. The wider usage and popularity of the indicator across a large geographical coverage (population) is the most important determinant factor of the strongest traditional EW indicator. Further to this, there are certain indicators that carry more weight than others depending on the source. For example, an EW delivered from traditional seer caries the biggest weight (it overrides all the others). Further interrogation about how one type of indicator can receive such a wide geographical acceptance coverage, confirmed that this is normally controlled by undocumented standard which is mostly based on the family linage of a certain seer (24which is passed from one family member to another). However, reliance on these traditional seers is slowly declining.

24

a family of seers can decide to settle in different parts of the county and by so doing ensure standardized or similar warning indicators for a particular prediction is disseminated across the large geographical area.e.g. cattle raids, drought etc

23

6.3. The Turkana Communities Through their traditional strategies, the Turkana community has been able to cope with harsh and unpredictable environment and to overcome the consequences of natural disasters. However, since the colonial intervention, their traditional system and way of life have increasingly been interfered with through a process of fragmentary modernization that threatens the viability of the pastoral strategies and makes the Turkana people more vulnerable to droughts. This underlines the understanding that livestock development interventions must be based on indigenous knowledge systems and on the traditional skills of these pastoralists to survive in their surroundings25. 6.4. Adaptive strategies of the Turkana community The herd owners in Turkana follow a number of conscious strategies which aim at coping with the hazards of pastoral life and survival in their arid and unpredictable environment. Some of these strategies include the manipulation of the natural resources through livestock management (species diversification, herd dispersal, herd maximization) and mobility; others refer to the building-up of social exchange networks and systems of risk-sharing. Turkana pastoralism is based on 5 animal species: camels, cattle, goats, sheep and donkeys. This diversification strategy stimulates a more complete use of the vegetation resources, it serves the different productive roles of livestock, and it reduces the risks of hard depletion in times of disaster. The domestic stock species do not all have the same food habits. Camels are mainly browsers while cattle graze. Sheep and goats form an intermediate group. The Turkana seasonal calendar is based on 6 months of wet and 6 months of dry period. Although, the Turkana communities have their own traditional EW indicators, they expressed their frustration caused by the changing weather pattern which has made it extremely difficulty for them to predict seasons as they used to do in the past.

6.5. Traditonal EW indicators for drought The following are traditional Drought Early Warning indicators among the Turkana community. The indicators can be grouped in the following categories;

25

Gert de Gans, 1986.

24

Early Warning (Human Nature and behavior) -

-

-

The traditional seers would foretell season through interpretation of their dreams – they would in the following morning request the community members to donate an animal for sacrifice. This animal would normally come from the most humble clan in the community. They believed an animal for sacrifice from such a clan/person was the most acceptable to god. Observation/studying intestines of a shoat to foretell looming dry/ bad season by Foretellers (seers) in the community. This doubles to foretell possible raids from cross border communities and forewarn the communities to put in place mitigation/prevention measures to averse the attack. When communities are involved with many social ceremonies e.g. marriages etc, the Turkana community believe this is a pre-caser for drought.

Early Warning (Animal behaviors) -

-

-

-

-

Young goat kid suckling while lying down or crouching- sign of a bad season/ drought or a young goat kid putting a one foreleg around the head is assign of drought – the severity of the drought is further emphasized by how far the kid places the foreleg round the head. When livestock starts to feed on dry bones and droppings of other animals – (the Turkana community perceive this to mean that the situation is so bad that the animals forced to feed on their own fresh) and this symbolizes a terrible upcoming drought that will clear the whole herd. When animals refuse to obey to follow the directions the herd man is gathering them to follow and run way scuttling in different/opposite direction symbolizes bad omen or drought When after it has rained and pasture in the fields is good and for some unknown reason the livestock is not fattening- this is indicator of an upcoming drought. When livestock become highly infested by animal pests especially tick and lice - this is a sign of drought on the way. When the Turkana herders witness an abnormal increase of shoats calving twins- this is an early indictor of drought Outbreak of livestock diseases indicate bad season/drought

Early Warning (Meteorological indicators) - Absence of rains and pasture drying out - Drying of surface water collection points like water pans and wells is sing of drought - Misty and high speed dusty windy storms symbolize drought

25

Early warning (migration of birds) -

Birds‟ migration – when certain birds (Elelee) soar very high in the sky is an indicator that the rains are far (drought), the opposite is true, when they fly low it gives an indication that the rains are approaching.

Early Warning (Astronomical) -

Movement and location of the stars ( constellation) The color of the moon -brown/dark symbolizes a bad omen/drought When the new moon rises, the shape of the Moon and tilting position eg: symmetrical shape (Lotiak) – symbolizes No rain that month. When a new moon rises and the crescent is directed to the North (left) which is known locally as Lodunge – is an indication of a bad season/ sign of drought. If on the other hand the Crescent is Tilting right - this symbolizes sign of wet season/rains and therefore the coming season is hoped to bring good rains.

6.6. EW indictors from 5 Turkana communities generated in a PCVA26 The following are indicators generated during PCVA meetings. Name of Community CMDRR site

Makutano Community (Turkana West District)

Lopusiki Community (Turkana West District)

Traditional drought Early Warning Indicators -

Observation & interpretation of shoats‟ intestines. Prediction by traditional foreseers/Healers Mist Kids suckling while lying down Livestock feeding on bones Stars constellation

-

Examination and interpretation of shoats‟ intestines. Multiple/ increased breeding in livestock. Water volume decrease and pasture drying. Prolonged dry season ( beyond six months) Young ones suckling while lying down. Livestock feeding on bones Observation of shoats‟ intestines Prolonged dry season Increased breeding in shoats (multiple kidding)

26

This compares well with my findings from different communities in the same county and further underlines the importance and weight of some of the key indicators that seem to be adhered to by the whole Turkana community.

26

Kataboi Community (Turkana North District- Lake region)

Lomekwi Community (Turkana North District- Lake region)

-

Kids sucking while lying down Livestock excreting while lying down. Shortage of fish in lake Turkana Traditional foreseers Migration of bird- when certain birds (Elelee) soar high, rains are far. Out breaks of diseases during wet season. Crescent Moon shapes

-

Decrease in water and pasture drying Livestock not fattening up during the rainy season. Livestock tend to return to the sheds when let out to graze. Poor catch when fishing Birds‟ migration – (Elelee) when they are high in the sky (drought) , when they soar low (rains are approaching) Foreseers When young ones suckle while lying down Crescent Moon shapes. Multiple kidding Constellations  Ng‟akanyer – group of stars that revolves the earth  Ekaal (Camel) – when it moves from its normal position, it is said that there is drought. Kids placing a foreleg on the neck  Short extension from the neck- drought will be less severe  Long extension from the neck- drought will be more severe Observation of shoats‟ intestines Pasture drying and reduced water volume. Seasonal calendar (six months wet and six months dry) Foreseers High temperatures during the day. Increased breeding Young ones suckling while lying down. Kids placing a foreleg on the neck  Short extension from the neck- drought will be less severe  Long extension from the neck- drought will be more severe. Grasshoppers/ Locust invasion

-

-

Riokomor Community (Turkana North District- Pastoralists)

-

-

Adapted from OXFAM Lodwar

The above indicators were collected during community risk analysis workshops meant to develop community disaster management plans. Most of the indicators are the same and thus widely used in the community.

27

7.0. TRADITIONAL EWS CHANNELS TO INFORM DISTRICT LEVEL ANALYSIS The traditional drought EWS of the two communities in Turkana and Wajir are based on three precepts. First pastoralists have detailed knowledge of when the major rains should arrive; - understanding what is implied about the probability of future rain by variance in wind, humidity and temperature from expected conditions. Second, the pastoralists know how to interpret the behavior of animals and plants, which serve as valuable indicators for slight fluctuations in temperature and humidity. When this is appropriately interpreted, local communities can forecast major rains four weeks before their arrival. Similarly, the absence of these natural indicators suggests the absence rains i.e. drought. Thirdly, the pastoralists observe historical trends that provide opportunity for reasonable predictions of future weather patterns. However, the increasing severity and frequency of drought over the last decade has rendered this latter form of forecasting less reliable than it has been the past. The indigenous communities in the two study areas recognize that drought is a an integral component of their environment, and for them to be able to cope with it , they have to monitor its patterns, make relevant predictions and choose appropriate actions and mitigation options based on traditional EW knowledge inherited from their ancestors. Lack of documentation of this traditional EW knowledge which is held by elders in the two communities however, is gradually disappearing before it is passed to the next generation. As mentioned earlier, both the Somali and the Turkana elders attributed this to several reasons; key to this is modern civilization and system of education that have no affirmative consideration for communities depending entirely on pastoralism for livelihood27.

7.1 Recognizing the value of traditional EW in the scientific EWS The deliberate lack of recognition and undermining of traditional community elders (who are the custodian of traditional DEW) by local government administrators have also contributed to the erosion of traditional EW knowledge The concerns expressed by the elders is collaborated by Platt (2002: 64) ……. the education curriculum told in Kenyan schools could indirectly be undermining the availability of labor force within the pastoral communities since the education framework does not incorporate pastoralism. Consequently, when the youth from the (pastoral communities) graduate from school, they are unable to effectively undertake pastoralism as a source of livelihood and earn a living (unlike their counter parts from Agricultural farming regions in the country). 27

28

and there is need for urgent documentation of and recognition of this knowledge. Among the Somali community in Wajir, there are strong religious believes that forecasting/prediction is a sin and against the Koran and this could be a key contributing factor in discouraging the generational transfer of the traditional knowledge on EW from elders to the next generation. This was more particularly outspoken in Wajir South. Thus , most community members steered clear from sharing their traditional knowledge for fear this might sent the wrong impression that one was „predicting the future‟, a act prohibited in the „Koran„. However, some community members and some key informers interviewed clarified this perception and rightly argued that, most of the indicators are not predictions, but in fact weather forecasts- based on seasonal observations of weather patterns and the behavior of animals that logically precede precipitation. Interestingly, the traditional rain indictors provide information that is no different than data obtained via satellites and atmospheric observation stations that evaluate changes in humidity, barometric pressure, and wind speed over time28. Though not quantitative, the conclusions obtained from community traditional knowledge are not far different from those disseminated through the media or by KMD as weather forecast, after scientific data have been compared against historical trends. The representatives from ALRMPs from the two areas were also in agreement that both the traditional community EWS and the scientific EWS have similarities, however, there is no clear official link between ALRMP‟s EWS and traditional EWSs from the two communities despite the fact that their field monitors also collect traditional EW information from the communities to enrich the monthly EWB. This information is only appreciated and not recognized or clearly indicated in the documentations as a contribution from traditional EW knowledge and wisdom of the respective communities. „This mode of drought monthly EW monitoring data collection exercise is not participatory and is more extractive than participative and in some cases loosing credibility due to inadequate supervision of field monitors‟29 Interestingly, the KMD staff manning the two field stations in both Wajir & Turkana appreciates the traditional EW knowledge held by communities although they confirm it is not included or captured in the weather reports. It would be interesting to see local weather outlook being aired over the local FM 28 29

Rebecca Huss-Ashmore, u.d; Platt, 2002: 64. Gerhard van‘t Land & Mike Wekesa: 2008.

29

radio stations. The KMD can be challenged to start considering county based daily weather outlook- briefs through local FM radio stations, especially in view of the proposed devolved services in the new constitution. County specific weather outlooks will be exciting and relevant to local communities.

7.2 Dissemination of EW Initially, traditional EW knowledge is a preserve for men (based on the general assumption that women are marginalized in all decision making structures in the two pastoral communities). However, it was evident that women are well conversant in traditional drought EW indicators especially those related with livestock behavior. First, this could be explained from their traditional roles of interacting with livestock (in both Turkana & Wajir communities, women are traditionally responsible for milking of animals). Secondly, and significantly so, the personal concern about the welfare of their families and particularly children during drought period could strongly explain this high level of awareness about TDEW indicators among women. Unfortunately, the high illiteracy level among the women poses a serious impediment and challenge in understanding of the DEW information disseminated through the ALRMP monthly EWB. This calls for a deliberate choice of a second alternative and appropriate channel for dissemination of EW information especially to grass root communities for effective impact.

The intention of the monthly drought EWB produced by the ALRMP in the two counties is to disseminate EW information on looming famine/droughts to relevant stakeholders and communities. However, the general feeling on the ground is that this bulletin primarily serves only two categories of stakeholders. One, it serves decision makers (at both the district and National levels) to forewarn them of potential food shortages that in the absence of intervention could lead to starvation. Second, the bulletin sends signals to donors of potential appeal for humanitarian support. The intention of ALRMP is to however to disseminate this information to the community level. Again the facilitation of EWS feedback information to communities is weak and sometimes the modes of getting the conventional EW information by communities from the DMO are either by obtaining the monthly EWB or physically visiting the DMO offices and seeking assistance.

30

7.3. Channels for enhancing EWS, using TEWS and scientific EWS -

Inclusion of Traditional EW information in EWB - Although the traditional EW opinions from the communities are consulted by ALRMP monitors, their views are not incorporated into the EWB generated by the ALRMP. There is need to review the packaging of the current EWB to include the quantitative traditional EW information collected from the communities. This will stimulate increased uptake of the EWB information by the communities. Oxfam & other stakeholders should therefore lobby ALRMP to redesign the EWB and include a provision for traditional EW input from the communities in their monthly bulletins.

-

Future exit strategy for ALRMP- At the moment, the ALRMP which is responsible for EWS is heavily depended on donor support. It should be in the interest of the responsible Ministry to start figuring out the future sustainability of this programme – particularly the EWS component. One way of getting out of this, would be commencing of gradual transfer of the EWS to local community institutions. This can start by designing and adopting a model tool for gathering information on the TEW indicators and analyzing them which will lead to production of TEW bulletin whose information can be blended with ALRMP bulletin for decision making. This process will in future provide an exit strategy for ALRMP, thus giving communities more responsibilities to gather and sound the alarm as the drought situation sets in. The District Pastoral Associations can be tasked with the TEW on behalf of the ALRMP. It was reported, in Wajir for example, the DPA have attempted to engage in this process in the past.

-

Simplify EW information to grassroots communities- The study established that the EWB serves as an important tool for drought contingency planning and management. Currently this is being utilized by GOK ministries and NGOs with offices and presence at the district level. However this information hardly reaches the grassroots communities. Again, the EWB is relatively scientific and given the high illiteracy levels among the communities in the two areas, there is need to simplify this information and explore a suitable dissemination methodology that is appropriate to the target communities.

-

Limitation of EWB- Currently the ALRMP dissemination of DEW information is done through the distribution of the EWB. This method is unsuitable, especially for mobile nomadic families, besides the high illiteracy levels and interpretation challenges posed by the technical content of the EWB. Further, the poor road infrastructure makes it extremely impossible to adequately cover the vast counties. Rather, the ALRMP should explore other

31

means to sound the warning and advice the communities on what action they need to undertake to avert the imminent disaster or cushion them against the damaging effects before a response is triggered. Use of local FM radio stations can assist in enhancing the uptake of EW information contained in the EWB. -

Increase uptake of EW information - The KMD weather forecasting in the two counties is largely received through the transistor radio (which is the only available media for local communities). Unfortunately, communities interviewed expressed very little faith with this information. This is a very worrying situation and a serious breakdown in any EWS. To change this situation, there is urgent need to facilitate ALRMP & KMD staff (based at the district level) to hold regular field community dialogue meetings. This will increase the confidence of the communities and enhance the lost credibility of the information they (ALRMP & KMD) gather and disseminate. The dialogue meetings will provide good opportunity for communities to learn and become aware about the EW & weather forecasting process of the two agencies, possible complementary opportunities and vice versa. This will scale up EW information uptake by communities and trigger consequent early actions (EAs).

-

Autonomy of the EWS - There is no doubt that EWS is the thread that links all the sectors of development and serves as a catalytic tool for reducing risks of disasters in the two counties. However, in its present state, its visibility has been significantly overshadowed by the implementation of the hardware drought management components within the ALRMP. In order to make the EWS more interactive with the communities, it might be crucial to explore the possibility of making the EWS more autonomous from these other (hardware) components of ALRMP; such as the drought management, support to local development and the Natural Resource Management (NRM). This way, the early warning information can get out of the offices in good time and reach the communities without necessarily being slowed down by other processes within the ALRMP.

-

Using FM radio stations to disseminate hybrid EW information- The study established that there is an array of FM radio stations in the 2 counties and Oxfam & other stakeholders should explore the possibility of using these FM radio stations to disseminate EW information to the communities. It is important to note that any initiative in this direction should adapt hybrid EW information system (traditional EWS meshed with scientific EWS) to enhance community confidence on drought EW indicators and consequently increase the uptake of EW information.

32

-

Strategic partnerships in dissemination of EW information- for sustainability purposes of broadcasting and dissemination of EW information, stakeholders should explore the opportunity of possible strategic partnerships. For example, in Turkana, the Diocese of Lodwar (DOL) owns FM radio station and Oxfam & other stakeholders can explore the possibility of partnering with DOL & ALRMP to air radio DEW messages. Again, there is already a radio programme broadcasting weather related information in Turkana Language aired through the KBC Kisumu station. But unfortunately this does not include traditional EW information.

-

A channel for generation of hybrid EWS using DRR sites as Convergence points for TEWS & Scientific EWS -Already Oxfam field staff have facilitated 26 communities to generate Community Disaster Management plans (CDMPs) through a risk analysis process. Oxfam should explore the possibility of facilitating the establishment of a sustainable hybrid EWS using these sites as focal points in the two counties. This can be done by linking the 26 DRR community groups with the ALRMP drought field monitors. Through the CDMPs, the DRR committees can help collect TEW indicator information for the monitors. These DRR groups will then become focal points for convergence of both the TEWS & Scientific EWS at the community level - and by extension through the ALRMP field monitors be linked to the ALRMP EW analysis centers located at the District level.

-

Community Awareness on Linkages between EW & Drought Contingency activities-Communities are very much aware about the ALRMP hardware contingency activities/interventions – but cannot connect these activities with EWB – this could be explained by high illiteracy levels and lack of linkages between the two – hence the need to improve linkages between the EW information and the drought contingency plans generated ALRMP.

-

Verification of early warning data - There were concerns raised by line ministries that all early warning reports should be verified (at the district level) before dissemination to make sure that they do indeed represent the trends of the events on the ground. This begs for an oversight data verification committee probably composed of departmental heads/ representatives of line ministries who can work closely with DMOs to ensure that the data collected by field monitors is a fair observation and professional interpretation of the specific sectors.

33

8.0. RADIO MESSAGES FOR COMMUNITY LEVEL EWS More often than not, EW information fails to reach the targeted groups due to several factors. These includes complexity of warning indicators, lack of proper analysis, complexity of analytical data, use of inappropriate language in relation to the end users and lack of communication channels or simply lack of institutions to facilitate dissemination of the EW information As earlier analyzed, the consultant established that the most trusted traditional EW indicators from both communities (Turkana and Somali) are connected with domestic animals30. Besides this, we argue that, indicators related to pasture and vegetation (which livestock depend on) are intertwined with behavioral indicators of the domestic animals and therefore the early warning messages should be generated around these two spheres of indicators . These indicators are often cited as the most dependable and as such, most likely to set in motion proactive drought mitigation actions by the two communities. Again culturally speaking, pastoralists have a strong attachment to livestock and it would be safe to generate EW radio messages around livestock. Also, the conditions of both vegetation and pasture have clear satellite generated indicators that can be linked with migratory nature of the pastoralists. When the pasture is depleted, it is most likely the water points are also running out. Early warning messages should also relate to the mobility of pastoralists. Mobility is one of the most successful strategies used by the two communities to deal with uncertainty and risk of drought. By maintaining a high degree of mobility, they take their animals to “follow the rains” and thus the availability of forage. Mobility also allows them to avoid pests (e.g. ticks) and diseases that arise when conditions become wetter. This shifting may (sometimes) serve as a means to avoid conflict with other groups using the same resources for part of the year. For the Turkana, this lifestyle also reduces the risk of becoming a victim of cross border livestock raiding. In addition, mobility helps in conserving biodiversity.31

The close proximity and interaction of human with livestock explains why most of the indicators relating to drought come from the observed behavior of domesticated animals, particularly camels and cattle’. (Platt: 2000: 26). 30

In Turkana, many pastoralists in more recent times, started to use mobile phones to help them manage mobility (e.g. finding out about conditions in other areas and also to have easier and quicker access to market information. This modern adaptive strategy was also noted by many scholars of this subjects (Yohannes & Mebratu 2009). 31

34

8.1. Area specific Radio messages 8.1.1. Wajir -The radio messages for DEW based on Wajir seasonal calendar: Season Jilaal – Jan to March

Guu- April to June

Annual Seasonal Indicators

Radio Message

High temperatures, Weathering of trees and shading of leaves , Drying out of water pans & Environment becoming dusty, Livestock production reduces, Increased water consumption, Clear clouds & Cracks in the grounds

Emergency period Emergency livestock off take to get money to purchase food and water. Information on livestock markets Hardest period, drought emergency response plans eg. stock pile fuel for borehole machines , repair of water pumps etc

High temperatures at the beginning of April, Nimbus clouds cover the sky, Very cold at nights, Trees produce buds, High dusty winds – Toloolia, Anxiety of animals (animals become „jovial„) – Beeyriiq, Reports of straying of camels, Mating of Shoats – Qalaat, SahaanHerders move out in search of livestock pasture,& Break out of common colds

Guu bula kalaheys, sac iyo salah ulahaw” Announce starting of the Guu or long rains. Good season if rains are not depressed. Rain geographical coverage and Start returning livestock back home from dry grazing areas.

Hagaay – July Too windy through out the season – Shurey, There is rain Drizzles, Livestock skin change – There is out break of livestock & human to September

Deyr – Oct to Dec Short rains

Cathaar, Reduced animal weight, Out break of diseases – Shimbir, hergeb, Strong wind that dries the grass pasture, Reduced pasture, Malnourished children & Sheep experience running noses and mucus High temperatures – Kaliil, Unpredictable clouds, There is anxiety among , community members as uncertainty creeps in, Reports of animals getting lost, strong will to survive among community , members -people not giving up, Trees producing buds and wild fruits – Mareer - Howlan , & NGOs start collecting Data from the community( if the short rains fail there is uncertainty of a long drought period.

diseases, announce type of diseases and how to get treatment drugs. Name and prize of livestock drugs

If Short rains are good- Prepare for bumper harvest.(Deyr bula kalaheys, rati ii rii ulahaw). If the short rains fail, prepare for along drought period. Migrate livestock to dry season grazing areas, and provide information on livestock migratory routes, start putting in place community drought contingency plans, Sale selected and healthy animals for maximum returns, provide information on livestock markets.

35

8.1.2. Turkana The Radio messages for DEW based on Turkana seasonal calendar Month of the year November

Turkana name of Seasonal calendar

December

Nakamu

February

March April Nakiparo

June

Livestock migration, No milk production, surface water dries up, community mining water from dry river beds, Trees shed leaves and fruits

Announcements on save dry season grazing areas ,

Cloud formation, light showers, plenty of rains, pasture growth, flowering & sprouting of under growth plants, No migration marriage ceremonies & cerebrations

Normal.

Dry season , pastures changes color, planning for migration

Drought Alert

July

August September

Radio messages

Emergency stage

Jan

May

Annual seasonal Indicators

Nait

Encourage pastoralists to sell healthy mature animals for maximum returns. announcement on drought contingency plans

Returning back home from dry season grazing areas Community awareness education on outbreaks of water born diseases during the ran season

Announcements on contingency plans & information

36

drought update

8.1.3. General DEW radio short messages for both Turkana & Wajir The following is the proposed suitable format for long term weather forecasting and Dissemination of DEW radio short messages for pastoralists in Wajir & Turkana Time scale Years

Example of Early Warning Increasing risks of severe drought due to climate change.

-

Months

Forecast of below average rainfall for the coming season. This is collaborated by TEW indicators

-

High frequency of water pumpsover usage by pastoralists

-

Water tankaring

-

Weeks

Low water tables in boreholes, drying of surface water sources eg. Water pans, depletion of pasture and vegetation cover. Forecast of depressed rains leading to high probability of rains failure and drought. This is collaborated by TEW indicators from the community. -

-

Days

-

-

Revise drought contingency plans, Sale selected animals when still in good condition. Announce prevailing food prizes in the market(Livestock/food commodities prize variations inform communities about drought enhanced risks and what to do if the risks materializes e.g. Advice pastoralists to stock pile dry food stocks ; Start gradual livestock movement & relocation (follow pasture & water) etc Announce livestock off-take programs and agencies involved. announce drought situation surveys announce prevailing market prizes of livestock and key food commodities Prepare for migration to dry season grazing arrears. Announce drought assessment surveys Announce relief intervention measures and agencies involved. Announce criteria for selection of food aid beneficiaries Announce Water tankeing collection points for both human consumption and livestock.

Water tankeing for both human consumption and livestock continues More Pastoralists settlements along main roads All water well levels extremely low Total failure of seasonal rains.

-

-

Hours

Example of FM Radio early action Announcements Continually upscale DRR and climate change adaptive measures. Advocate for a mixture of modern & appropriate traditional CC adaptive livestock management strategies.

All animals have migrated and

-

Commence total migration to dry season grazing areas. Announce critical relevant information e.g safe passage corridors, water points along the corridor, security situation (Turkana) and destination areas. announce relief food intervention measures by various agencies Announce final relief distribution points and anticipated food aid rations Announce Arrival of relief trucks at final food distribution

37

vulnerable family members (children, the elderly, women, school going children) are left behind

-

points. Announce number of targeted beneficiaries in each area.

8.2. Proposed Hybrid Radio EWS Dissemination process The study proposes the piloting of an hybrid DEWS which will be a mix of the scientific EW (generated by ALRMP monitors and KMD weather forecasters ) and the TEWS input (getting monitors to interview DRR committees) then the two sources feed into ALRMP EWB. The final product is then disseminated through stakeholders and FM radio stations for airing to communities. Fig 1 & 2 are schematic presentations of the proposed hybrid EWS. Fig.1. Hybrid EWS

38

Fig 2 proposed hybrid EWS

The ARLMP District office will still remain the focal point for processing and generation of EW. Here the scientific early warning information from KMD weather forecast, the data gathered by the ALRMP field monitors and the traditional early warning data generated by communities through the DRR committees will be converged. After analysis and verification (by line departments), the Hybrid Early Warning generated will then be packaged for different stakeholders. A summary of simple and clear EW messages will first be sent via FM radio stations to communities. This will be followed by sharing a detailed EWB with DSG members and other stakeholders at both district and National levels for their planning purposes Care should be taken to ensure that the proposed hybrid EWS is not viewed as a parallel EW for drought or coming to replace the exiting ALRMP EWB. Instead the HEWS should be seen as the normal scientific EWS, but this time round with input from TEWS. This will not only strengthen the reception and confidence of communities about DEWs information, but will also increase the uptake of these warnings to trigger timely response by pastoralists. Moreover, the proposed channel of dissemination of HEWS -through the local FM vernacular radio stations will increase geographical coverage and timely EWs.

39

To advocate and lobby for the adoption of the hybrid EWS, Oxfam should target the following key stakeholders at 3 levels: i. National level-The Drought Management Initiative, The ALRMP Head Quarters, Kenya Meteorological department (KMD), ICPAC, The Ministry of Special programmes ,The Ministry of State for Development of Northern Kenya, UN OCHA , the Kenya Food Security Steering Group(KFSSG) etc ii. District Level-The DSGs, DMOs , the Practical Action, the DOL32 (Turkana), Oxfam partners in Wajir33 (WASDA, DPA & ALDEF), 34strategic NGOs, KMD field station staff, local FM radio stations, District Pastoralist Associations, District Livestock Marketing Committee etc iii. Community Level- DRR committees, Community opinion leaders, local Administration e.g. Dos & chiefs and ALRMP Field Monitors

9.0. CONCLUSION There is no doubt that, the drought frequency is increasing in both severity and length. This is attributed to climate change and variability among other causes. In the prevailing situation, any potential low cost strategy that can improve the survival and mitigation of drought related losses exposed to these pastoralists‟ communities should not be ignored. This therefore calls for a deliberate strategy to seek the „best of both worlds „ by exploring how some of the best and strongest traditional drought EW indicators and the best scientific EW indicators (KMD plus ALRMP) can be combined to offer a local hybrid inclusive drought warning system that is simple, culturally trusted by the pastoralists. To address the current dismal uptake of DEW information and geographical coverage by the monthly EWB produced by ALRM, the proposed hybrid EW information could be disseminated to communities through the local FM radio stations and monthly community assemblies (meeting) in DRR sites especially in 32

The (DOL) Diocese of Lodwar has an FM radio satiation already in Turkana- could be a suitable partner in dissemination of HBEW especially if one chooses to ride on the large networks of the church in Turkana county. The radio is covering about 100km radius at the moment. 33 the OGB partners in Wajir will play an important role for long term sustainability of the HBEWS in Wajir county 34 These are NGOs who have previous experience in setting up traditional EWS especially on Kenya/Uganda cross boarder livestock movements – these include Practical Action & VSF Belgium in Turkana.

40

areas where the FM radio stations are limited in coverage because of the vastness of the Counties. This will guarantee wider geographical coverage-to the advantage of the nomadic pastoral families. With solar powered transistor radios, mobile families can monitor the advances of a potential drought and make informed decisions based on timely EWs messages.

41

REFERENCES Action Aid. 2006. Risk Reduction. Human Policy Briefing note 2. (Online). Retrieved from: http://www.actionaid.org/assets/pdf%5disaster risk reduction-note.pdf. (2011, May 13). Ashmore – Huss Rebecca;(ud): Local Level Data for use as Early Warning Indicators; Department of Anthropology ,University of Pennsylvania, Philadelphia. Sommer Florian, 1998: Pastoralism & Drought Early Warning Response . Baumwell Jennifer, 2008: The Value of indigenous Knowledge for Disaster Risk Reduction; A unique Assessment tool for reducing Vulnerability to Natural Disasters. Webster University, Viena. C. Pratt. 2001. Traditional early warning systems and coping strategies for drought among pastoralist communities. Fletcher School of Law and Diplomacy, Tufts University, Medford, MA, United States. Environmental and Societal Impacts Group. 2004. Early Warning systems and sustainable development. ESIG AlertNo.6. (Online). Retrieved . from http//www.isse.ucar.educ/alerts6.html. (2011, May 17). Gerhard vaan‟t & Mike Wekesa. 2008: Functional Analysis of Drought Management at District Level; Study for the EC funded Drought Management Initiative(DMI) in collaboration with the Arid Lands Resource Management . Howell Philippa, 2003: Indigenous Eaerly Warning indicators of cyclones. Potential application in Coastal Bangaladesh. Banifield Hazard Research Center. Hab,S. Asch.K., Farnandez-Steeger,T.,& Arnhart,C.2009. People centered landslide early warning systems in the context of risk management. Geographical research abstracts VO. II. EGU200911466-2. Hanover , Germany : EGU General Assembly. (Online) Retrieved from : http//www. meetingorganizer .copernicus.org/EGU20009-11466-2.pdf( 2011, May 18). Monnik,K.2000.Role of drought warning systems in South Africa‟s evolving drought policy. http//www. drought .unl.edu/monitor/EWS/ch5_ Monnik.pdf( 2011, May 14). Platt Christopher, 2002: Traditional Early Warning Systems and Coping Strategies for Drought Among the Pastoralists Communities Northeastern, Kenya. Fletcher School of Law & Diplomacy. Samarajiva,R & Waidyanatha,N. 2007. Early Warning Systems in disaster risk reduction . IDRCCIDA workshop: Moratuwa,Sri Lanka, 112 July 2007(Online). Retrieved from: http//www.lirneasia.net/wp-content/uploads/2009/samarajvaews12july07.ppt(2011, May 12). Stepheson, R.S.1994a.Disaster assessment. 2nd edition.UNDMTP.s.l;Artifax. UNISDR, 2006a . Developing Early Warning systems. Wilhite, D.A & Svoboda, M.D. 2000: Drought early warning systems in the context of drought preparedness and drought management .

42