Directorate of Resource Surveys and Remote

Sensing (DRSRS)

The use of Earth Observation data and emerging technologies for

Forest Monitoring:
Landsat, NICFI, Planet, Sentinel, MODIS-VIRS and Jazamiti App in
Kenya.
Merceline Ojwala, DRSRS - Presented during COP28 in Dubai: 09-12-2023

Collaborators: GMA, AWAK,

TNC, University of
Leicester, Cambodia and
DRSRS Sumitomo Corp. (Japan)
 Land use Transitions
Comparison of change in maps from 2 time
periods sequentially (e.g. 2002 vs 2006, 2006 vs
2010, 2010 vs 2014, and 2014 vs 2018).
Change map with;
Areas remaining in the same land cover type
Areas changed to different land cover type
between 2-time periods for the specific
REDD+ activities.

Sustainable Forest
No-change Enhancement Degradation Afforestation Deforestation Non-forest Transitions
Management
 Methods of Data Acquisition Through EO
SAMPL POP POP POP STD.
SPECIESNAMES E SUM ESTIMATES VARIANCE ERROR

Cattle 43,293 571,280 2,087,295,267 45,686

Donkey 763 10,068 5,462,362 2,337
Sheep & Goat 89,446 1,180,299 5,479,647,330 74,024
Buffalo 865 11,414 22,060,018 4,696
Eland 197 2,599 842,407 917

Multi-Stage Sampling Concept

Elephant 116 1,530 245,165 495
Grants Gazelle 198 2,612 412,253 642
ThomsonsGazelle 1,825 24,082 36,111,898 6,009
Giraffe 308 4,064 1,469,490 1,212
Impala 1,325 17,484 12,980,504 3,602

Stage 1: Remote Sensing Approach Kongoni

Ostrich
Topi
56
41
947
738
541
12,496
181,254
37,302
7,016,919
425
193
2,648

 Orbiting Space Satellite (3,000 - 35,000 km)

Warthog 137 1,807 172,815 415
Waterbuck 2 26 641 25
Wildebeest 10,358 136,680 1,134,455,619 33,681
Zebra Burchells 4,954 65,371 180,700,500 13,442

Vertical Aerial Photography(100-3,000m) Pop. imate

Est
180000
167000
1600
00

Hyperspectral remote sensing

y = 10569x-1.15
1200 R² =0.806
00

4000 35462
0
2000 21573 15 168 177
13139
0 80 00 02
1
0

LidarApplication
19 1977-80 1981-85 1986-881989-91 1992 2000
73 Year -94 -04

Stage 2: Aerial Surveys

 Low-High FlightAircraft
Aerial Photography Animal Census (100-
200m)
Costs Implication: Dependent on size of
area, sampling resolution and efforts
OUTPUTS
 Maps
Stage 3: Ground Surveys/Measurement  Statistics
 Models
 Attribute identification, scale  Reports
accuracy and socio-economic surveys
Cost Implication: Often expensive and time • Planning,
• conservation,
consuming
• Management
• Policy formulation
• Sample unit 70m x 70m (~0.5ha)

• 49 internal grid points (15%)

• Sample unit = interpretation + spatial

support unit

• Planet Data (NICFI)-Monthly composites,

2016 – 2021, 5m resolution

• Sentinel data - Monthly composites, 2015

– 2022, 10m resolution

• GEE- Landsat and Sentinel composites,

Time Series data

• Google Earth Pro - Very high resolution

data

• Geo Dash - NDVI TS graphs, Normalized

Difference Fraction Index, Landsat
Composites
 MONITORING NATIONAL TREE
GROWING IN KENYA
 One of H.E. the President’s directive is to have 15 billion
trees grown in the country to increase tree cover from the
current 12% to 30% by 2032.

 DRSRS is hosting JAZAMITI App which is a multi-agency

innovation that leverages on ICT to document, verify,
monitor and report tree planting and growing activities
across all the 47 counties in Kenya.

 This initiative is to solve the gap of Monitoring,

verification and reporting of trees that are planted across
the country to establish survival rate.
Tree cover Monitoring
JazaMiti App
9

Thank You!