GRASS as a Temporal GIS - Sören Gebbert

IntroductionTime in GRASS GIS

Temporal Modules

GRASS as Temporal GIS

Sören Gebbert

THÜNEN Institute of Climate-Smart Agriculture

November 23, 2013

Sören Gebbert 2013 GRASS as Temporal GIS


Temporal Modules

Topics

1 Introduction

2 Time in GRASS GIS

3 Temporal Modules



Temporal Modules

GRASS GIS

GRASS Geographic Resources Analysis Support SystemMultipurpose raster, 3D raster and vector based GISDeveloped 1982 - 1995 at CERL in Illinois/USASince 1999 open source under GPL license with activeinternational development communityVery modular, provides GUI, command line and batchprocessing supportAbout 500 modules for management, processing, analysisand visualization of geographical data are available(GRASS version 7 2013 Aug.)



Temporal Modules

Raster processing

Slope, AspectExpositionspatial aggregationMap algebraSpline InterpolationPrincipal components...



Temporal Modules

Vector processing

Generalize, BufferPatch,Overlay,Select,ExtractShortest PathTraveling salesmanproblemDelaunay, Voronoitriangulation...



Temporal Modules

Why a temporal GIS?

We need to model and assess GHG emissions, Soil OrganicCarbon (SOC) change and Land Use Change (LUC) atcontinental scale with different spatial and temporal resolutions



Temporal Modules

Why reinventing the wheel?

Available temporal GIS and related multi purposeenvironmental modeling solutions do not fit our needs



Temporal Modules

Topics

1 Introduction

2 Time in GRASS GIS

3 Temporal Modules



Temporal Modules

Temporal GRASS GIS goal

Development of the first comprehensive field basedtemporal GISEfficient management, processing, analysis andvisualization of large spatio-temporal fields and theirinteractionsProviding interoperability between sophisticatedspatio-temporal analyzing tools like CDO, R and ParaViewCreating an intuitive object oriented spatio-temporalframework



Temporal Modules

What is a field based temporal GIS?

Two important Temporal GIS classifications (Goodchild 1989,Heuvelink 1998)

a) Temporal GIS dealing with objects like points, lines, arcs,areas that represent geometrical or topological features inspace and time

b) Temporal GIS dealing with fields, natural, social orepidemiological. Their attribute data representing thedistribution of temperature, precipitation, hydrological orecological patterns in space and time

We use the field based approach in GRASS GIS.



Temporal Modules

Temporal GRASS GIS concept

Do not break the existing GRASS GIS functionalities andassure backward compatibilityFollow the UNIX paradigm, create small modules for aspecific purpose and combine them to manage complextasksDesign time support for the existing datatypes: raster, 3Draster and vector map layersDesign new spatio-temporal datatypes, space timedatasets, that manage time series dataReuse existing raster, 3D raster and vector modules toprocess space time datasets



Temporal Modules




Temporal Modules

Temporal relationships

Supported temporal relationships (Claramunt and Jiang 2001).



Temporal Modules


Use interval time and time instances of absolute(Gregorian calendar) and relative time (years to seconds)Use a SQL database to store the temporal and spatialextent of space time datasets, map layers and theirmetadataUse the SQL database to store relations between mapsand space time datasetsImplement a comprehensive object oriented temporalframeworkImplement new GRASS modules based on the temporalframework for managing, processing and analyzing ofspace time datasets



Temporal Modules

Space time datasets



Temporal Modules

Space time datasets



Temporal Modules

Topics

1 Introduction

2 Time in GRASS GIS

3 Temporal Modules



Temporal Modules

Modules to manage space time datasets

t.createCreate a space time dataset

t.registerAssign time stamps and register raster, vector or voxel maplayers in a space time dataset

t.unregister

Unregister map layers from space time datasets

t.removeRemove space time datasets



Temporal Modules

t.register



Temporal Modules

Modules to manage space time datasets

t.supportModify the metadata of a space time dataset

t.topology

Shows and checks the temporal topology of a space timedataset

t.shiftTemporally shift a space time dataset

t.snapCreate a valid temporal topology of a space time dataset



Temporal Modules

Modules to process space time raster datasets

t.rast.listList registered raster map layers. Support SQL WHEREstatements as well as methods like: list by time order or list bygranularity

t.rast.seriesPerforms different aggregation algorithms on all or a subset ofraster map layers in a space time raster dataset

t.rast.extractExtract space time raster datasets from an existing STRDSusing SQL WHERE statements and map-calculationexpressions.



Temporal Modules

Modules to process space time raster datasets

t.rast.mapcalcSpatio-temporal raster algebra

t.rast.aggregateTemporally aggregate a space time raster dataset usingdifferent statistical aggregation methods

t.rast.univarCalculates univariate statistics for each registered raster maplayer of a space time raster dataset



Temporal Modules

t.rast.aggregate



Temporal Modules

t.info precipitation_1950_2011_yearly

+-------------------- Space Time Raster Dataset -----------------------------+| |+-------------------- Basic information -------------------------------------+| Id: ........................ precipitation_1950_2011_yearly@PERMANENT| Name: ...................... precipitation_1950_2011_yearly| Mapset: .................... PERMANENT| Creator: ................... soeren| Creation time: ............. 2013-09-18 13:35:16.243647| Temporal type: ............. absolute| Semantic type:.............. mean+-------------------- Absolute time -----------------------------------------+| Start time:................. 1950-01-01 00:00:00| End time:................... 2012-01-01 00:00:00| Granularity:................ 1 year| Temporal type of maps:...... interval+-------------------- Spatial extent ----------------------------------------+| North:...................... 75.5| South:...................... -0.5| East:.. .................... 75.5| West:....................... -40.5| Top:........................ 0.0| Bottom:..................... 0.0+-------------------- Metadata information ----------------------------------+



Temporal Modules

t.info precipitation_1950_2011_yearly

+-------------------- Metadata information ----------------------------------+| Raster register table:...... precipitation_1950_2011_yearly_PERMANENT_raster_register| North-South resolution min:. 1.0| North-South resolution max:. 1.0| East-west resolution min:... 1.0| East-west resolution max:... 1.0| Minimum value min:.......... 0.0| Minimum value max:.......... 2059.0| Maximum value min:.......... 16881.0| Maximum value max:.......... 35116.0| Number of registered maps:.. 62|| Title:| Yearly precipitation 1950 - 2011| Description:| Yearly precipitation 1950 - 2011 in [0.1 mm]| Command history:| # 2013-09-18 13:35:16| t.rast.aggregate| input="precipitation_1950_2011_monthly"| output="precipitation_1950_2011_yearly" base="precip_yearly"| granularity="1 year" method="sum"| # 2013-09-18 14:28:35| t.support in="precipitation_1950_2011_yearly"| title="Yearly precipitation 1950 - 2011"| descr="Yearly precipitation 1950 - 2011 in [0.1 mm]"|+----------------------------------------------------------------------------+



Temporal Modules

t.rast.aggregate + t.rast.extract + t.rast.univar + R

●●●●

●●

●

●●

●

●●●

●●

●

●●●●●●●●

●

●●

●●

●●●●

●

●●●

●

●●●●

●

●

●●

●

●●

●●

●●

●

●●

●●●●

●

1950 1960 1970 1980 1990 2000 2010

−5

05

1015

20

Years

Tem

pera

ture

in d

egre

e C

elsi

us

Seasonal mean temperature trend of the temperate European climate Zone from 1950 − 2010

● Summer temperature, linear regression slope 0.019Fall temperature, linear regression slope 0.006Spring temperature, linear regression slope 0.023Winter temperature, linear regression slope 0.019



Temporal Modules

Import and Export of space time raster datasets

t.rast.export and t.rast.import



Temporal Modules

Modules to process space time vector datasets

t.vect.listLists registered vector map layers of a space time vectordataset

t.vect.extractExtracts a subset of a space time vector dataset. SupportsSQL WHERE queries for metadata and attributes.

t.vect.univarCalculates univariate statistics of attributes for each registeredvector map layer of a space time vector dataset



Temporal Modules

Modules to process space time vector datasets

t.vect.db.selectPrints attribute data of vector map layers of a specific spacetime vector dataset

t.vect.observe.strdsObserve specific locations in a space time raster dataset over aperiod of time using vector points

t.vect.what.strdsStore values of raster map layers at spatial and temporalpositions of vector points as vector attributes



Temporal Modules

t.vect.observe.strds + t.vect.db.select + R

●●

●●

●

●

●

●●

●

●●

●

●●

●

●●●●●●●

●

●

●●

●●

●●

●

●●

●●

●

●

●●●●

●

●

●●

●

●

●

●

●●

●

●

●●

●

●●

●

●

1950 1960 1970 1980 1990 2000 2010

−5

05

1015

2025

Years

Tem

pera

ture

in d

egre

e C

elsi

us

Seasonal mean temperature trend of Berlin from 1950 − 2010

● Summer temperature, linear regression slope 0.023Fall temperature, linear regression slope 0.011Spring temperature, linear regression slope 0.034Winter temperature, linear regression slope 0.029



Temporal Modules

Import and Export of space time vector datasets

t.vect.export and t.vect.import



Temporal Modules

Modules to process space time raster 3D datasets

t.rast3d.listList registered 3D raster map layers.

t.rast3d.univarCalculates univariate statistics from the non-null cells for eachregistered 3D raster map layer of a space time raster 3Ddataset

t.rasedt.mapcalcSpatio-temporal raster 3D algebra

t.rast3d.extractExtract space time raster 3D datasets from an existingSTR3DS using SQL where and r.mapcalc queries.



Temporal Modules

Animation with g.gui.animate



Temporal Modules

Comparison of two maps with g.gui.mapswipe



Temporal Modules

Time-line visualization with g.gui.timeline



Temporal Modules

The End

Thank you


Date post:	10-May-2015
Category:	Technology
Upload:	luisdesousa
View:	388 times
Download:	2 times

GRASS as a Temporal GIS - Sören Gebbert

Technology