• No results found

Cloud-based Geospatial Data services and analysis

N/A
N/A
Protected

Academic year: 2021

Share "Cloud-based Geospatial Data services and analysis"

Copied!
47
0
0

Loading.... (view fulltext now)

Full text

(1)

Cloud-based Geospatial

Data services and

analysis

Xuezhi Wang

Scientific Data Center

Computer Network Information Center Chinese Academy of Sciences

(2)

Outlines

1

Introduction of Geospatial Data Cloud

2

Cloud-based Geospatial Data Analysis platform

3

Cloud-based Geospatial Data visualization

platform

(3)

Geospatial Data Cloud

I.

Platform overviews

II.

Geospatial data services

III.

Geospatial model services

IV.

Services achievements

(4)

Why do we do?

In 2008, USGS opened all geospatial datasets to the public for free.

• Landsat, MODIS is the most popular remote sensing dataset for scientists and researchers. Especially, the

Landsat series.

• Due to the low international network speed, many Chinese scientists cannot take advantage of these resources

easily.

• Based on the CNIC’s mass storage and high speed network( CSTNET: Chinese Science and Technology Network), we mirrored the geospatial datasets

covered China and surrounding area, and provided online data services for

(5)

Design a new cloud-based geospatial analysis and

visualization platform USGS opened its Landsat datasets.

we contacted the USGS staffs through email.

We got an oral permit for Landsat redistribution from USGS. we download the Landsat datasets, and provide public

services for Chinese scientists

Cooperated with National Center for Atmospheric Research (NCAR)

Mirrored the NCAR datasets provide public services for Chinese

Scientists

We got a permit for SRTM dataset redistribution from International Centre for Tropical Agriculture(CIAT). We cache

SRTM global elevation data, and provide online data services.

Accroding to the user requirements, We make some remote sensing products from Landsat dataset and Modis dataset.

such as vegetation index, water cover dataset.

Dowload MODIS dataset covered CN Host the first user conference Design a new cloud-based data platform

Geospatial data cloud opened for public services Host the second user conference with ESRI China

(6)

Platform objectives

Based on our cloud infrastructure, we hope to construct an 'one-stop'

geospatial data service platform for Chinese scientists and researchers.

An open cloud-based platform meeting the requirements of data

search, data acquisition, online storage, scientific models, online

calculation and map visualization.

Model Online Calculation Search Acquisition Storage Visualization

http://www.gscloud.cn

(7)

General framework

Database

cluster Online storage

Near-line storage Data integration Automatic Metadata crawling Automatic data entities cache Mass Storage Cloud environments Distributed file system Virtual Machines Task scheduling Data query Data category Integrated search Complex search One-key indexes Data services Smart data Order Data application

Favorite data Batch

download Model services Model package Online execution Web map

services Map publish

Map visualization Map style edit Model management User data spaces Data management Task management Map management Model flow Map share Service management System managem ent Users Data Models Indexes Tasks Services Resources monitor

(8)

Geospatial Data Cloud

I.

Platform overviews

II.

Geospatial data services

III.

Geospatial model services

IV.

Services achievements

(9)

Online data resources

 Open geospatial datasets: LANDSAT、MODIS、EO-1、DEM、NCAR、

NOAA and LUCC

 More than 350TB data entities, and over 7 million data records.

 Covered China and surrounding area

Online services for Chinese scientists and researchers with free of charge

Landsat1-3 MSS Landsat 4-5 MSS Landsat 4-5 TM Landsat 7 SLC-on Landsat7 SLC-off Landsat 8 OLI TERRA AQUA TERRA AQUA GDEM 30m SRTM 90m Ocean elevation HYP_L1R HYP_L1G AL1

(10)

Generated geospatial products

Water-cover of China from Landsat TM/ETM in Summer

Slope, Aspect, Topographic position indexes et. of China

Land Surface Temperature, Vegetation indexes (5 days, 15

days, and 30 days) of China, from 2000 – now

ASPECT

(11)

Geospatial data query

Massive geospatial metadata

I need Landsat data in summer, 2000

I need DEM data of Beijing

Input: geo-location or address

PostgreSQL+PostGIS Cluster Based on geospatial

storage cluster and search engine, we provide three ways for users to access their preferred data:

1: by address name 2: by geo-location

(12)

Platform Screenshots

(13)

Platform Screenshots

(14)

Platform Screenshots

(15)

Platform Screenshots

(16)

Platform Screenshots

The result page that matched the query criteria

Extra attribute

(17)

Geospatial Data Cloud

I.

Platform overviews

II.

Geospatial data services

III.

Geospatial model services

IV.

Services achievements

(18)

Geospatial model services

Why do we need this services?

For example:

If we need Beijing DEM data, there have two ways.

Traditional way: search DEM data, download DEM data,

mosaic DEM data with ArcGIS, and then clip DEM data

with Beijing shape mask .

Convenience way: select Beijing administrative unit in

the web, submit a clip task to system, and then download

Beijing DEM data.

Which is the best way?

I prefer the second way.

That’s why we provide online geospatial model

(19)

We developed an elastic online task processing system with cloud techniques.

Users fill model parameters on the web page and submit tasks to the system;

The tasks were run under cloud virtual machines;

The results were stored in cloud storage.

Task scheduling Task Task Task Computing

resource pool Storage pool

VM VM

VM

virtual machines in Cloud

Overall frameworks

(20)

Online models and tools

DEM tools:

DEM clip and mosaic; Reprojection;

Aspect, Slope, Topographical index calculation of

specified region

LANDSAT tools:

Online gap-fill model for Landsat ETM7 SLC-off

dataset

MODIS tools:

Online Land Surface Temperature calculation of user

specified time range

Online vegetation indexes calculation of user specified

time range

(21)

Private cloud storage

We provide user with private cloud storage to store

their own model inputs and outputs data.

User manage their own data here, including upload, download, rename, delete, create folder and move

(22)

Platform Screenshots

(23)

Platform Screenshots

Model parameters page of DEM clip and mosaic tool

(24)

Platform Screenshots

Task tracking page

list user’s submitted tasks

(25)

Platform Screenshots

Task result page

When the task was

finished, the results can be download from here, or from user private cloud storage

(26)

Geospatial Data Cloud

I.

Platform overviews

II.

Geospatial data services

III.

Geospatial model services

IV.

Services achievements

(27)

Our platform was wildly used in many research fields in China: earth

science, environmental science, ecology, hydrology, meteorology,

oceanography, natural disasters, city planning ...

(28)

Based on statistics in November 2013

User characteristics

Currently, over 70,000 users

Doctor Master Bachelor Others University CAS Government Company Others

(29)

Service statistics

Avg. downloaded over: 200 GB/day;

online downloaded over: 250 TB;

(30)

Outlines

1

Introduction of Geospatial Data Cloud

2

Cloud-based Geospatial Data Analysis platform

3

Cloud-based Geospatial Data visualization

platform

(31)

Cloud-based Geospatial Data

Analysis Platform

Why do we need this platform?

For example: If we want to calculate Chinese monthly

temperature variation from 2000 to 2014, how can I

achieve this?

Traditional way: download all data, and calculate with workstations. It

may spent several months to download data. This is a very bad thing.

Shortcuts: using cloud computing technology. Apply virtual machines

in the cloud; Write code in the virtual machines; read all data from

cloud storage; Run program in parallel

Which way is the best? I prefer the second way.

That’s why we want to construct this platform.

(32)

Overall frameworks

...

Storage resource pool

Cloud manager with

OpenStack

Distributed file system

Data pool

Computing resource pool

Public geospatial

data

User private

geospatial data

virtual machine pool

VM

VM

VM

(33)

Currently progress of the platform

We have finished the prototype system. The system is

under inner tests for access pressure and stability. In the

end of this year, it will be provided for the public access.

(34)

Outlines

1

Introduction of Geospatial Data Cloud

2

Cloud-based Geospatial Data Analysis platform

3

Cloud-based Geospatial Data visualization

platform

(35)

Cloud-based Geospatial Data

Visualization Platform

Why do we need this platform?

In many cases, users want to publish their geospatial data,

model results, vector map to the web site;

They also want to customize the map layer styles through

the web pages;

They also want to customize the components and styles of

the webgis pages;

(36)

Optimized Web Map Service

Computing

resource pool Storage pool

VM VM

VM

Dynamical render map in Cloud

HTML5 map client ... Tiles Storage Dynamical cache Map styles editor Map request

Overall frameworks

(37)

Currently progress of the platform

Most of system functions were already finished,

under inner tests and bug fixed.

Published services provided:

OGC WMS(web map service), Google TMS , OGC

WFS (web feature service)

(38)

System screenshots

I.

Render multi-bands raster layer with RGB false-color,

(39)

System screenshots

(40)

System screenshots

(41)

System screenshots

(42)

System screenshots

(43)

System screenshots

II.

Render vector layer with colorful symbol by unique

(44)

System screenshots

III.

Render vector layer with colorful symbol by classified

(45)

System screenshots

IV.

Render vector layer with sized symbols by classified

(46)

System screenshots

(47)

Thanks

Any suggestions?

References

Related documents

understanding of a number of the concepts included in the ICB.3 and in due course IPMA will develop supplementary exams to enable PRINCE2 Practitioners to achieve appropriate IPMA

Commercial Cloud Providers Vendor Instance storage Object storage Block storage Semi- structured data storage Relational Database storage Distributed File System Online

Under the background of nursing staff shortage and rapid population aging in China, the shortage of elderly care professionals is as severe as the situation of nurses.. Besides

The purpose of this project is to establish meaningful relationships between the Seventh-day Adventist Church and the Muslim community in Grand Cape Mount County through

With the aim of re-situating modern Iranian drama within the context of international dramatic developments, the plays of Gholamhossein Sa‘edi, Akbar Radi, and

Georgia Tech Athletic Association Gordon College Athletics Gwinnett Braves Baseball Club Hometown Sports Collectibles Jenkins County Recreation Lake Arrowhead Yacht Club

The main reasons given for bypassing PHC facilities in the studies mentioned above were poor services (includ- ing lack of drugs and diagnostic services) and lack of trust in

To start in hard boot mode, hold both the Mode button and the X100 button pressed in while powering on the fixture.. When the Mania SCX500 is in hard boot mode, the data