-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.Rmd
183 lines (112 loc) · 4.15 KB
/
index.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
---
title : Introduction to GEE
subtitle : Digital Soil Mapping
author : José Padarian
job : University of Sydney
framework : io2012 # {io2012, html5slides, shower, dzslides, ...}
highlighter : highlight.js # {highlight.js, prettify, highlight}
#hitheme : tomorrow #
widgets : [] # {mathjax, quiz, bootstrap}
mode : selfcontained # {standalone, draft}
knit : slidify::knit2slides
biglogo : usyd_logo.png
logo : usyd_logo_big.png
---
## Outline
> * What is GEE and how it works
> * Explore GEE API (__Javascript__ and ~~Python~~)
> * DSM workflow (with examples)
* Compile covariates
* Extract data from covariates
* Model training
* Generate map
* Export
--- .segue .dark .nobackground
## Google Earth Engine
--- .nobackground
<img src="assets/img/google-products.png" width="800" class='centered'>
---
## Features
* Access to over 40 years of Satellite Imagery (with daily updates)
* Algorithms to analyse that data
* Google's computer clusters to analyse the data at scale
* A web-based tool for interactive data exploration
--- .segue .dark .nobackground
## How does it work?
--- bg:#FFF .nobackground
<img src="assets/img/how_1.png" width="83%">
--- bg:#FFF .nobackground
<img src="assets/img/how_2.png" width="83%">
--- bg:#FFF .nobackground
<img src="assets/img/how_3.png" width="83%">
--- bg:#FFF .nobackground
<img src="assets/img/how_4.png" width="83%">
--- bg:#FFF .nobackground
<img src="assets/img/how_5.png" width="83%">
--- bg:#FFF .nobackground
<img src="assets/img/how_6.png" width="83%">
--- bg:#FFF .nobackground
<img src="assets/img/how_7.png" width="83%">
--- bg:#FFF .nobackground
<img src="assets/img/how_8.png" width="83%">
---
## Features
* Access to over 40 years of Satellite Imagery (with daily updates)
* Algorithms to analyse that data
* Google's computer clusters to analyse the data at scale
* A [web-based tool](https://code.earthengine.google.com/) for interactive data exploration
--- .segue .dark .nobackground
## DSM workflow
---
## Compile covariates
<img src="assets/img/covariates.png" class='centered' style='margin-top:50px;margin-bottom:50px'>
\> 200 public datasets
\> 4000 new images every day
\> 5 million images
\> 5 petabytes of data
---
## Examples
[Image and Image Collection](https://code.earthengine.google.com/97a8ece1d4ea58db05d5317e1e55ea80?noload=1)
[Basic calculations](https://code.earthengine.google.com/141d968b0b815625f1003ea215c3ef4f?noload=1)
---
## Extract data from covariates
* At the moment only using Fusion Tables
* Another alternative is using GeoJSON (usually for small datasets)
* Limit of 500,000 rows (also a size limit)
We will be using data [SOC from Edgeroi](assets/data/edgeroi_WGS84.csv)
# Let's import data into [<img src="assets/img/google_fusion_tables_logo.png" style="padding-top:20px">](https://drive.google.com/)
---
## Examples
[Feature Collection](https://code.earthengine.google.com/a970ea5bff3005119947380000a53112?noload=1)
[Extract covariates](https://code.earthengine.google.com/1ad8188d7f26cb631993de0548dd202b?noload=1)
---
## Model training
* Classification and Regression algorithms
* Limited number of algorithms available
* Tree-like (CART, Random Forest)
* Perceptron
* Support Vector Machine
* Others
* Alternative to use expressions (tree-like structures, linear models)
---
## Examples
[Train model](https://code.earthengine.google.com/b0f1030f127383b797e995b92b4d3515?noload=1)
---
## Generate map
This is the step where we can get most speed gain
* The predicion is made at the viewport
* It depends on the extent
* It depends on the zoom level (resolution)
It's not always necessary to view the prediction (lazy evaluation)
---
## Examples
[Apply model](https://code.earthengine.google.com/f06a57cc8638d542385e510182a0c811?noload=1)
---
## Export
There are multiple alternatives:
<img src="assets/img/google_drive_logo.png" style="height:125px">
<img src="assets/img/google_cloud_storage_logo.png" style="height:100px">
It is also possible to export images as Assets
---
## Examples
[Export](https://code.earthengine.google.com/bf3a7ba541cf8c07cccbe2806fa80af5?noload=1)