The document discusses how new technologies like satellites, unmanned aerial vehicles, and internet of things can help improve agricultural statistics collection, especially for smallholder farms. Satellites provide high-resolution optical and radar imagery that can monitor fields and crops. UAVs allow real-time, on-demand data collection that is not constrained by clouds. IoT enables automatic, around-the-clock sensor data collection. These technologies help provide statistics on crop areas, yields, growing seasons, and cropping patterns. The document also discusses several initiatives using satellite data and machine learning to map crop types and estimate yields for smallholders in places like South Africa and Ethiopia.
Liangzhi You (IFPRI) • 2021 IFPRI Egypt Seminar Series: "Fostering Digitalization for a Future-Proof Food System in Egypt"
1. Digitalization and Food System:
Statistics From Space
Liangzhi You
International Food Policy Research Institute (IFPRI)
Webinar on “Fostering Digitalization for a future-proof food system in Egypt”
January 26, 2021
5. Newtechnology
UnmannedAerialVehicles(UAVs)
What it is
Captures super-high-resolution remote
sensing data on the fly.
What’s new
In-field analysis of UAV-captured imagery
within minutes (not days any more).
Relevance for ag stats
Allows on-demand real-time monitoring of
smallholders’ plots. Not constrained by
clouds.
What it takes
Technical partner (equipment, piloting),
capacity for imagery analysis and
interpretation, regulation compliancy.
https://www.pix4d.com
6. Newtechnology
Internetof Things(IoT)
What it is
A system of inter-connected computing devices
that generate data and transfer.
What’s new
Solar-powered hyperlocal, real-time sensing of
weather, water accounting, and canopy
management data.
Relevance for ag stats
Automatic and systemic collection of in-field
groundtruth data 24/7/365.
What it takes
Technical partnership for the installation and
management of sensor network.
https://www.arable.com
8. • Cropland (areas)
Permanent (areas)
Arable (areas)
Rainfed/irrigated (system areas)
Growing seasons (periods)
Cropping patterns (areas)
Crop specific (areas and yields)
Hierarchy of Relevant Crop-related Statistics That Might Be
Improved Using Earth Science Data
13. The Land Use Change Knowledge Integration
Network (LUCKINet):developgriddeddataproductsthatdepictthedynamicsin
variousimportantland-usevariablesoverthelast~25years
18. Smallholder croptypeand yield estimationusing satellite dataand
machinelearning approach
• New advances in AI offer promise for smallholder crop
area and yield estimation
o Publicly available satellite data - Sentinel(10m, 5-day revisit)
o Free cloud computing platform (Google Earth Engine)
o No/low cost of (deep) learning packages
• Crop type mapping in South Africa
o Deploy TensorFlow, an open source deep learning platform with time
series of sentinel data for crop type mapping.
o Limited ground truth samples (<500 samples per crop)
o Distinguished major crop types including inter-cropping and fallow land
in Free State in South Africa with R-squared of 0.71
• Crop yield estimation in Ethiopia
o The deep learning neural networks outperform other machine learning
algorithms.
o vegetation index from Satellite + climate variables + soil give the best
model performance
o The maize yield estimates has R-squared of 0.62 across three AEZ
zones.
For more information, please contact Joe Guo (z.guo@cgiar.org)
Fallow
Maize
Pasture
SoyaBeans
Sunflower
Vegetables
WheatMaize
WheatSoya
Non crop area
Reference
Predicted
Predicted
yield
Reference yield
Digitization changes every aspects of agriculture, from production, precision agriculture, market, value chain to agricultural R&D (e.g. gene technology, big data and machine learning in agronomy), digital agricultural extension, digital finance. That also raise question on digital divide, digital inclusion. I’ll pick up only one aspect: how digitization revolutionizes traditional agricultural production information system, specifically agricultural statistics.
RDD has two branches: The Census and Survey Research Branch and the Geospatial Information Branch
Within GIB developing methodologies and tools to improve NASS’ ability to collect, manage, and disseminate statistics on US agriculture utilizing remotely sensed, GIS, and GPS data
Skill set composed of:
Geographers
Cartographers
Information Technology
Programmers
Statisticians