global -scale analysis of earth observation data
DESCRIPTION
The Google Earth Engine. Global -scale analysis of Earth observation data. Dave Thau [email protected]. Who are we?. - Technology-Driven Philanthropy. Goal: Use Google's strengths in information and technology to address global challenges. Google’s Strengths. - PowerPoint PPT PresentationTRANSCRIPT
Google Earth Engine - [email protected] - TWDG – September 28, 2010
Who are we?
- Technology-Driven Philanthropy.
Goal: Use Google's strengths in information and technology to address global challenges.
2
Google Earth Engine - [email protected] - TWDG – September 28, 20103
Google’s StrengthsData management, processing, Storage, indexing & delivery.Making things simple, at scale.
Global ChallengesClimate changePandemic diseasePoverty
RE<C and RechargeITDeveloping utility scalerenewable energy cheaper than coal (RE<C) and accelerating the commercialization of plug-in vehicles (RechargeIT)
Google Power MeterGoogle PowerMeter is a home energy monitoringtool that gives you theinformation you need touse less electricity andsave money.
Google Flu TrendsGoogle Flu Trends uses aggregated Google search data to estimate flu activity in near real-time in 20 countries.
3
Google Earth Engine - [email protected] - TWDG – September 28, 2010
Motivation
UNEP: "Atlas of our Changing Environment"
4
Rondonia, BrazilJune 1975
Rondonia, BrazilJuly 1989
Rondonia, BrazilSeptember 2001
Google Earth Engine - [email protected] - TWDG – September 28, 2010
Challenges and GoalsAccess to Datao Cost, availability & licensing of petabyteso Raw data requires preprocessing Why should each user have to solve this on their own?
Access to Software, Algorithms, and Hardwareo What algorithms are available and where to find them?o I don’t have a cluster, what do I do?
Non-technical Challengeso Open, transparent, verifiableo Privacy and sovereignty concerns
Non-Goals• Doing global monitoring ourselves• Scientific algorithm research
5
Google Earth Engine - [email protected] - TWDG – September 28, 20106
The Plan
Developing a platform for processing at scale• Inherently parallel system• Instant visualization• Promote transparency, reproducibility, reuse
Getting expertise from the experts• Avoid reinventing the (scientific) wheel• Partner with academic institutions, NGOs, indigenous people…
Providing a place to publish (or not)• Let users curate their own datasets and algorithms• Help others to find them• Or keep everything private
6
Google Earth Engine - [email protected] - TWDG – September 28, 20107
Google Earth Engine - [email protected] - TWDG – September 28, 20108
Google Earth Engine - [email protected] - TWDG – September 28, 20109
Google Earth Engine - [email protected] - TWDG – September 28, 201010
Google Earth Engine - [email protected] - TWDG – September 28, 201011
Google Earth Engine - [email protected] - TWDG – September 28, 201012
Google Earth Engine - [email protected] - TWDG – September 28, 201013Demo
Google Earth Engine - [email protected] - TWDG – September 28, 201014
Google Earth Engine - [email protected] - TWDG – September 28, 2010
Very fast computation of scientific map products• Using arbitrary (user-supplied) algorithms
With some nice features • APIs for algorithm development and Web front ends• Access control, versioning, provenance• Online and Desktop versions (open source desktop version)• TIFF / KML / Fusion Tables / Google Maps
On a lot of data• Stored in their native projection• Every available Landsat scene• Every MODIS scene• Commercial datasets with cost pass-thru• User supplied data
Details
15
Google Earth Engine - [email protected] - TWDG – September 28, 201016
Android phones+OpenDataKit Training: Tanzania Village Forest Monitors, TZ Forestry and Beekeeping Division, Jane Goodall Institute, Nov 2009
Google Earth Engine - [email protected] - TWDG – September 28, 201017
Each point: GPS-tagged photo, tree info, plot info, more….
Google Earth Engine - [email protected] - TWDG – September 28, 201018
…more data collected for each in-situ sample point
Google Earth Engine - [email protected] - TWDG – September 28, 201019
Dataset fromBrianHeidorn
Image from AnimalBiodiversityWeb
Google Earth Engine - [email protected] - TWDG – September 28, 2010
Niche Modeling
River health
Your Ideas
Biodiversity and Conservation
20
Prediction for Triatoma barberiPeterson, et al. 2002
Upper MississippiFrom USGS
Google Earth Engine - [email protected] - TWDG – September 28, 2010
Thank youTo sign up for our trusted tester list:
21