Communica)ng  downscaled,  probabilis)c  seasonal  forecasts  and  evalua)ng  their  impact  on  farmers’  management  of  climate  risks:  

Examples  from  Kaffrine  (Senegal)    and  Wote  (Kenya)  

Ousmane  Ndiaye    –  ANACIM  K.P.C.  Rao  –  ICRISAT  

Jim  Hansen  –  CCAFS,  IRI  Arame  Tall  –  CCAFS,  ICRISAT  

 

Hypothesis Since   many   farm   management   decisions   are  taken   without   knowing   what   the   season   going  to   be,   advance   informaHon   about   the   possible  seasonal  condiHons  will  help   farmers   in  making  more  informed  decisions.  

Sahel: Annual Precipitation

200

250

300

350

400

450

500

550

600

650

700

1900 1920 1940 1960 1980 2000

Rainfa

ll (mm)

Observed

Key constraints addressed •  Lack  of  awareness  about  seasonal  climate  forecasts  and  their  reliability  

•  MispercepHons  about  the  climate  and  its  variability  

•  Lack  of  understanding  about  the  probabilisHc  nature  of  forecast  informaHon  

•  Non-­‐availability  of  informaHon  in  a  format  that  can  easily  be  understood  by  the  farmers  

•  Dialogue  between  users  and  producers  of  climate  informaHon  

NaHonal  insHtuHons  working  on  food  security  (+  social,  disseminaHon)  

Local  expert  group  

Rural  radio   SMS  

Farmers    

Face  to  face  

PR

OD

UC

TIO

N

TA

ILO

RIN

G

CO

MM

UN

ICA

TIO

N

STEP 1: BUILDING AN INTEGRATED FRAMEWORK: THE MULTI-DISCIPLINARY WORKING GROUP

Seasonal  forecast  ⇒  varie)es  Onset  forecast  ⇒  farm    prepara)on  

Nowcas)ng  ⇒  flooding  saving  life  (thunder)  Daily  forecast  ⇒  use  of  fer)lizer  /  pes)cide  Decade  forecast    ⇒  weeding,  field  work  

Evalua)on  Lessons  drawn  

Training  workshop  Indigenous  knowledge  Discussion  and  mee)ngs  

Field  Visits  experts  mee)ng  each  10  days  :    monitoring  the  season  

Decade  forecast  ⇒  op)mum  harves)ng  period    Daily  forecast  ⇒  saving  crops  leS  outside    

Before   During  the  Crop  season   Maturity/end  

Methods used in Kaffrine (West Africa) and Wote (East Africa)

•  The  study  was  conducted  in  Kaffrine  disctrict  (Senegal)  and  Wote  division,  Makueni  district,  Eastern  province  (Kenya)  during  the  2011  &  2012  rainy  seasons  

•  Study  treatments  include    – Survey  (Control)  –  InterpreHng  and  presenHng  seasonal  forecast  informaHon  in  the  form  of  an  agro-­‐advisory  

– Training  workshop  along  with  advisory  – EvaluaHon  

Building  on  local  knowledge:  High  humidity  and  high  temperatures  can  explain  some  of  their  indicators  è  “Stronger  monsoon”  Doing  quite  the  same  thing  BUT  Be\er  observing  system  More  reliable  storage  capacity  (numbers,  maps,  computers,  …)  

« When the wind change direction to fetch the rain »

= Wind change from harmatan to monsoon during onset

STEP 2: BUILDING TRUST LINKAGE TO INDIGENEOUS KNOWLEDGE

team work : farmers, climatologist, World Vision, Agriculture expert, sociologist

“KNOWLEDGE SHOULD PRECEDE ACTION” Farmer in kaffrine

Wote: Observed responses

Treatment  Area  cul)vated  (ha)   Investment  

(Ksh/ha)  Yield  (kg/ha)  

PS   ES  

Control  (T1)   1.53   2.06   1797   386.8  Training  workshop  (T2)  

2.00   1.89   2043   447.3  

Agro-­‐advisory  (T3)  

2.04   1.62   6092   613.8  

Training  workshop  and  advisory  (T4)  

2.10   1.94   3400   441.4  

Expectation for the season

Village/treatment  Women  farmers   Men  farmers   All  

No   Yes   No   Yes   No   Yes  

Control  (T1)   82   18   82   18   82   18  

Training  workshop  (T2)   63   38   54   46   59   41  

Agro-­‐advisory  (T3)   53   47   42   58   52   48  

Training  workshop  and  advisory  (T4)  

27   73   33   67   30   70  

Ø  First  step  :  building  trust    (social  dimension  :  using  indigeneous  knowledge)  

Ø  Giving  not  only  useful  BUT  useable  forecast  (tailored  for  specific  user  needs)  

Ø  Long  term  and  mulH-­‐stakeholders  partnership  (each  insHtuHon  has  part  of  the  soluHon  for  food  security)  

Ø  CommunicaHng  probabilisHc  aspect  of  the  forecast  (easy  to  understand,  can  translate  into  acHon  and  to  evaluate)  

Ø  Dynamic  process  :  need  to  be\er  understand  farmers  decision  system  (long  term  dynamical  partnership)  

Ø  The  forecast  covers  a  large  area  :  we  need  forecast  at  farm  level  Ø  Farmers  sHll  lack  of  tools  and  materials  beside  climate  informaHon  

LESSONS AND CHALLENGES

Ø   « We  were  guessing  now  we  have  decision  tools  »  Ø   « The  early  warning  system  of  an  very  early  rainfall  saved  all  my  crops  lea  outsides»  

Ø   « with  eminent  rainfall  forecast  through  sms  (nowcasHng)  we  can  saveguard  our  ca\le,  return  from  farms  to  avoid  thunder  »  

Ø   « we  woman  (soeur  unies  de  Ngodiba)  are  now  be\er  of  and  as  equipped  as  men  now.  »

FARMER TESTIMONIALS (Kaffrine)

Demand for climate services (Wote)

Village/treatment  

Amount  willing  to  pay  (Ksh/season)  

Women   Men   All  

Training  workshop  (T2)   258   357   313  

Agro-­‐advisory  (T3)   228   204   211  

Training  workshop  and  advisory  (T4)  

385   364   368  

All  villages   262   263   261  

Methods  •  In  Kaffrine:  300  farmers  trained,  more  than  1000s  received  climate  services  (33%  of  women)  

•  In  Wote:  A  total  of  117  farmers  (61%  women)  accessed  and  used  climate  agro-­‐advisories  

•  Farmer  use  of  climate  informaHon  was  assessed  by  conducHng  three  surveys  –  Before  training  or  providing  forecast  informaHon  – During  the  season  – Aaer    the  season  

ACHIEVEMENTS

THANK YOU