You are on page 1of 33

What you see is what you believe

• Demonstrate the value of VHRI to
help scale up a few quick-win
productivity enhancement
technologies in 6 smallholder
communities across Burkina Faso,
Ghana, Mali and Niger

– Focus on the last 8 km

– Candidate technologies:
spatially optimized soil and
water management practices
– Show a variety of value-added
– Demonstrate real-world
deployability and potential
– Upload GIS datasets to shared
online AgCommons repository
– Publish metadata on
Phases, tasks
• 1: Organize project resources:
human, methods, tools
• VHRIbox
• VHRIex2
• 2: Roll out VHR and technology
related information at site level
• Roll out VHRI to farmer
fields (ROLLout)
• 3: Deploy VHRI incubators on-farm
and in-silico
• Populate VHRIbox through
in-situ interactions
• Explore VHRI support
functions for contour
ridge tillage (inSilico)
• 4: Package field data, lab data and
synthesize two-way feedbacks
• Collate information from
incubators, forward
updated maps to sites
and collect feedback
Very High Resolution? Veritably,
a Huge Revolution !!
VHRI containers
• Preliminary VHRI extraction, in
laboratory – no field experience
• Results to be evaluated against
data collected during onFarm phase
• 6 file geodatabases created with
base imagery (BASEIMG), biomass
proxies (BIOMASS), field limits and
land use (FIELDSLU), topography
and soil information (TOPOSOIL),
and ancillary data (OTHERS)
• 10 students trained in ESRI
geodatabase management
• Operator on-screen digitization
approach can be subjective
• DigitalGlobe copyright restrictions
⇒ QuickBird base imagery provided
in JPG panchromatic and
multispectral color composites only.
Full spatial resolution retained.
A Manifesto Against Top-Down
Take part? Be part.
• Importance of participatory

• Learning by doing, Learning by

• Listening and observing? Giant time
• Gender-aware team

• Farmer-led participatory sampling

approach: farmers (rather than
researchers) initiate group
formation (image group, non-image
group) themselves based on i/
family lineages, ii/ hamlets
distribution across the landscape,
iii/ relative size of hamlets in
number of households. This is an
intrinsically spatial, equitable
sampling protocol owned by the
Sowing the Seeds of the
Bigger Picture
ROLLout entry points
• Population: growth
• Constraint: increasing demand
on the production resource
base ⇒ GO LOCAL
• Opportunity: increasing
diversification of income
sources for livelihoods ⇒ GO
• Identify site-specific
technologies (and issues) that
can benefit from VHRI support
• VHRI as a tool to concurrently
• 6 sites, 3 days/site
• 44,000 km-person road travel
• Direct interactions with 183
• 600+ photos, 600+mn of
streaming video
Growing $trong
Tree carbon
• Originally: hotspots/brightspots
of soil fertility activity, but
trees in fields affect NDVI
• New objective: quantify tree
layer with carbon markets
• Circular Hough Transform (CHT)
can be optimized for tree
density estimation from
panchromatic VHRI
• CHT needs to be re-optimized
for each site and for new
• Individual crown size estimates
from CHT are more prone to
• Crown height from geometric
Rain or Drain…
That is the Question!!
Contour Ridge Tillage
• CRT specifically adapted to
Sudanian climate and
intensifying systems (PIS, SER)
• But, field sizes & geometry also
come into play (NOB)
• VHRI (+ satellite-derived DEMs)
can be used to infer slope
aspect, monitor CRT adoption &
provide some elements of
guidance for implementation on
a landscape scale
• However, satellite-derived
DEMs alone are not precise
enough for CRT installation at
the field level
• VHRI captures effects of CRT
on canopy establishment: more
biomass early (+69% avg in
77% of fields, -11% avg. in
rest), more heterogeneity early,
and more homogeneity late
Home Run with the Minister
of Agriculture
Public awareness,
capacity building
• 7 bilingual blog entries online
• Minister of Agriculture, Mali
• Parliament representatives for
Banamba, Diema districts, Mali
• Bilingual flyers produced, 14 x
500 (3 x 500 more coming up)
• 4 sets of Trimble® GPS and
DELL computer equipment
distributed to partner NARS
• 10 students gained experience
with VHRI analysis and
exchange with farmers
• Cross-fertilization with BMZ-
funded CODEWA project
Learnings: pluses, minuses
• Pluses:
– VHRI a fantastic tool for discussion support, and with demonstrated potential for a
range of decision support (agro-forestry, water harvesting, fertility management)
– a tool that helps both reveal (capture) farmer knowledge (including intrinsic spatial
skills), and bridge shortcomings
– enlightening interactions with farmers, learning opportunity for students, policy
makers clearly see the potential for rural land tenure
– AMEDD (Mali) is independently working with VHRI for land tenure negotiations in
rural areas (forthcoming blog/flyer)

• Minuses:
– 1-year timeframe more reasonable than 6-months
– 70K budget a limitation (field work logistics in widely distant locations)
– Scarcity of qualified students (for laboratory and field activities), limited availability
of experienced scientists for backstopping of field work
– more work is needed to evolve knowledge from VHRI-triggered information
exchange (from a science perspective: need of socio-economic backstopping), OR
MAYBE NOT (just dump the whole raw information without further delay)
Next steps and recommendations
• Now what?
– Complete knowledge extraction from onFarm data
– Serve data online on – including picture and video data

• Next?
– Continue to stimulate demand from locally elected officials through showcasing, with support
from AMEDD (Mali) – Banamba, Diema districts, National Assembly.
– Work out with AGCommons a communication strategy at a regional level
– Develop a larger R4D proposal to scale up approach to 12 other sites in West Africa. Identify
partners and questions in East Africa (SIBEA).
– Explore VHRI dissemination options with and without support. With suggested productivity
enhancement technologies, and without.
– Translate VHRI into national languages – interactive maps
– Delegate processing tasks. Explore crowdsourcing options:
• VHRIbox: online student crowdsourcing with open-source software (e.g. GoogleMaps)
• VHRIex2, ROLLout, FEEDfwd: through local NGO networks (to be identified)
• onFarm: offline crowdsourcing with student interns (e.g. APEJ in Mali – need to engage
political actors), embed VHRI in other participatory research
• inSilico: “re-discover” locally adapted technologies with VHRI