non-members

92 Pages·2014·2.85 MB·English

PE&RS
October 2014 Volume 80, Number 10
y
g
o
ol
n
h
c
e
d t
n
a
e
c
n
e
ci
s
n
o
ati
m
or
nf
al i
ati
p
s
o
e
g
d
n
a
g
n
gi
a
m
or i
al f
n
ur
o
al j
ci
offi
e
h
T
G
N
SI
N
E
S
E
T
O
M
E
R
&
G
N
RI
E
E
N
GI
N
E
C
RI
T
E
M
M
A
R
G
O
T
O
H
P
INTRODUCING THE ASPRS SMARTPHONE APP
The American Society for Photogrammetry
and Remote Sensing (ASPRS) has launched
an association and conference focused
smartphone app for use by ASPRS members
and those interested in the Society. The app
keeps ASPRS competitive and continues
to provide members with the tools they
find essential, such as real-time access
to association news, member directories,
information on ASPRS webinars and
conferences, and critical industry alerts all from
their iPhone or Android smartphone. The app is
available on both iPhone and Android platforms.
Now everyone can have the association The Conference side of the app allows Downloading the new ASPRS app is
experience right at their fingertips. The attendees to connect with the conference simple! iPhone users can visit the Apple
technology available through the smart- web page, access a dynamic electronic App Store and download the ASPRS
phone app includes accessing the ASPRS final program, create a personalized App free of charge. Android users may
mobile website, direct access to “My calendar of events during the conference access via the Google Plays Store to
ASPRS” portals with editing and log-in week, view a map of activities, restau- download the new version, also free of
capability, a student information tab rants and attractions surrounding the charge. Individuals may also find the
with focused activities and information, conference location, post personal experi- links to download the smartphone app
publications tab for browsing the ASPRS ences and leave live real-time comments on the ASPRS website. Download the
Bookstore and online publications, an for other attendees on the “Fan Wall”, a ASPRS App today!
upcoming webinar and events calendar “Socialize” tab for access to ASPRS social
and membership “Join Now” material. media pages, and an “FAQ” tab with the
most important questions and answers
In addition, the ASPRS smartphone app asked during the conference.
includes a second screen dedicated to
conference activities. The latest confer- ASPRS is continuously working to im-
ence information is uploaded to the app prove communications with our members
about a month prior to the upcoming and stakeholders, and expanding the mo-
conference and then requires the well- bile app beyond its initial conference-
known “shake to update” for the latest focused purpose in order to address a wid-
app information. er range of ASPRS programs and infor-
mation sources was a natural transition.
PECORA 19
Sustaining Land Imaging...
UAS to Satellites

in conjunction with the Joint Symposium of ISPRS Technical
Commission I and IAG Commission 4
November 17-20 The 19th William T. Pecora Memorial Remote Sensing Symposium – Sustaining
Land Imaging: UAS to Satellites will be held in conjunction with the Joint
2014
Symposium of the International Society for Photogrammetry and Remote
Sensing (ISPRS) Technical Commission I and International Association of
Geodesy (IAG) Commission 4. The Pecora 19 Symposium will be held November
Renaissance
17-20, 2014 at the Renaissance Denver Hotel, Denver, Colorado and will be
Denver Hotel
combining various general sessions and special technical sessions throughout
Denver, Colorado the week with the ISPRS/IAG Symposium.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING October 2014 913
The ASPRS Student Advisory Council (SAC) …
…ensuring Student Member representation in ASPRS and more.
SAC is a group of students committed to serving all of the student members
of ASPRS. Our goal is to ensure that ASPRS is a Society that both benefits
from student involvement and creates opportunities for those students.
SAC is led by a Council of seven students who meet monthly to discuss
issues pertaining to ASPRS Student Members. What do they do?
• Organize special sessions of interest to students at ASPRS Annual
and fall conferences. http://www.asprs.org/Annual-Conferences/
Program/
• Create networking opportunities during those conferences and bring together students looking for employment after graduation
with potential employers in the industry.
• Inaugurate new programs within ASPRS.
• Design activities such as the GeoLeague Competition where students compete in teams using geospatial technology applications
to solve a problem. http://www.asprs.org/Students/GeoLeague-Challenge-2014.html.
Promote student involvement in humanitarian projects such as crowdsourcing the manual interpretation
of imagery in Somalia to identify shelters that are being used as homes by refugees. http://irevolution.net/
tag/tomnod/.
All ASPRS Student Members are encouraged to become involved with SAC. Check out the SAC Social
Networking sites and keep up with ongoing news.
Student Newsletter: http://asprssignature.blogspot.com/
Facebook page: https://www.facebook.com/pages/ASPRS-Student-Advisory-Council/117943608233122
LinkedIn Group: http://www.linkedin.com/groups?home=&gid=2487675&trk=anet_ug_hm
Email: [email protected]
STUDENT ADVISORY COUNCIL
THE IMAGING & GEOSPATIAL
INFORMATION SOCIETY
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
The official journal for imaging and geospatial information science and technology
October 2014 Volume 80 Number 10
COLUMNS HIGHLIGHT ARTICLE
918 Detection of North American Land Cover Change
Letter from Chandra P. Giri, USGS 916
Industry Insight—Letter to the Editor 917 Between 2005 and 2010with 250m MODIS Data
Grids and Datums—Republic of Armenia 927 René R Colditz, Darren Pouliot, Ricardo M. Llamas, Collin Homer, Rasim Latifovic, Rainer
Mapping Matters 929 A. Ressl, Carmen Meneses Tovar, Arturo Victoria Hernández, and Karen Richardson
Behind the Scenes 931
PROFESSIONAL INSIGHT
In Memoriam—Joel Nevin O’Neal 932
Book Review—Remote Sensing of Energy 933 925 An Interview with Clifford J. Mugnier
Fluxes and Soil Moisture Content
ANNOUNCEMENTS
918 925
Pecora 19 & ISPRS Commission I Symposium 913
ASPRS Scholarships 936
ASPRS Seeks New Book Review Editor 937
Call for Papers 994
DEPARTMENTS
Certification 930
Region News 935
ASPRS News 936
New Members 937
Classifieds 937
Member Champions 938
Ad Index 938
Calendar 962
Forthcoming Articles 962
Who’s Who in ASPRS 995
Sustaining Members 996
Instructions for Authors 998 PEER-REVIEWED ARTICLES
Membership Application 1000 939 A Semiautomatic Extraction of Antarctic Lake Features Using Worldview-2 Imagery
Shridhar D. Jawak and Alvarinho J. Luis
This month’s cover
image shows the land The design and implementation of a novel semiautomatic method to extract lake
cover change of North features in cryospheric environments from remotely-sensed images.
America between the
years 2005 and 2010 as 953 Integration of Lidar and Landsat to Estimate Forest Canopy Cover in Coastal British Columbia
mapped with 19 classes
Oumer S. Ahmed, Steven E. Franklin, and Michael A. Wulder
by the North American
Land Change Monitoring The potential of Landsat imagery to accurately estimate forest canopy cover measured
System (NALCMS). The from small-footprint airborne lidar data was examined in order to expand the lidar
background shows the
measurements to a larger area.
land cover map of 2010
draped over the globe, 963 Object-Based Hyperspectral Classification of Urban Areas by Using Marker-Based Hierarchical
centered at 10°N and Segmentation
100°W, and complemented with imagery of different
sources for areas not mapped by NALCMS. The Davood Akbari, Abdolreza Safari, and Saeid Homayouni
close-ups indicate changes between the years 2005
A novel framework based on marker-based hierarchical segmentation for object-based
and 2010 for Lake Manicouagan in Quebec, Canada
classification of hyperspectral imagery using both spectral and spatial information.
(top) with different stages of vegetation succession,
forest loss due to a large fire in the Big Sur area in 971 Automatic Smoke Detection in MODIS Satellite Data Based on K-means Clustering and Fisher
California, USA (middle), and different mean annual Linear Discrimination
water levels of the Vicente Guerrero reservoir in
Tamaulipas, Mexico (bottom). Xiaolian Li, Jing Wang, Weiguo Song, Jian Ma, Luciano Telesca, and Yongming Zhang
For more information on the NALCMS project
An automatic smoke detection algorithm by integrating K-means Clustering and Fisher
including the legend, see the highlight article in this
issue. All land cover maps and change products can be Linear Discrimination in MODIS data.
dcweoecwb.onsrilgtoe aasdn, eded .fg fo.o rUr i nnthditeiev deid nSuttiarale tc ecosou:n nhtttirntipee:ns/ /tf lrfaornomdm ci onhvstettpirt.:u/t/iwonwawl . 9S8o3e EWxa. mMinyiinngt ,C Jhaanngeet FDreatencktiloinn ,A Mppircohaacehleas Bfour eTnroepmicaanl nM, aanngdro Wveo Mn oKn.i tKoriimng APPPLAICPAERTIONS
usgs.gov and Mexico http://www.conabio.gob.mx/
Using Landsat imagery acquired over select sites in three tropical
informacion/gis/. Data assimilation: René R. Colditz
countries (Bangladesh, Burma, and Thailand), this study employed different change
([email protected]), design: Bernardo
Terroba ([email protected]). We detection approaches and composite bands to determine the most effective approach to
kindly acknowledge the Commission for Environmental monitor changes in mangrove forests.
Cooperation (CEC) for covering all costs.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING October 2014 915
PHOTOGRAMMETRIC ENGINEERING
& REMOTE SENSING
LETTER FROM CHANDRA P. GIRI,
UNITED STATES GEOLOGICAL
Journal Staff
SURVEY Publisher Dr. Michael Hauck
Editor Russell G. Congalton
Technical Editor Michael S. Renslow
Assistant Editor Jie Shan
Dear Readers, Assistant Director — Publications Rae Kelley
Electronic Publications Manager/Graphic Artist Matthew Austin
Land cover mapping and monitoring is one of
the oldest and most popular applications of Photogrammetric Engineering & Remote Sensing is the official journal
remote sensing. This is partly because timely of the American Society for Photogrammetry and Remote Sensing. It is
devoted to the exchange of ideas and information about the applications of
and accurate information on land cover and land photogrammetry, remote sensing, and geographic information systems. The
cover change is needed for many and diverse technical activities of the Society are conducted through the following Technical
applications at local, national, continental, and Divisions: Geographic Information Systems, Photogrammetric Applications,
Lidar, Primary Data Acquisition, Professional Practice, and Remote Sensing
global levels. Although significant advancements Applications. Additional information on the functioning of the Technical
have been made in producing better spatial and Divisions and the Society can be found in the Yearbook issue of PE&RS.
thematic land cover data sets, quantifying the Correspondence relating to all business and editorial matters pertaining to
land cover change on a regular basis remains a this and other Society publications should be directed to the American Society
challenge. An additional challenge is to monitor for Photogrammetry and Remote Sensing, 5410 Grosvenor Lane, Suite 210,
Bethesda, Maryland 20814-2144, including inquiries, memberships, sub-
land cover change at a continental scale. The scriptions, changes in address, manuscripts for publication, advertising, back
highlight article of this issue, “Detection of North American land cover issues, and publications. The telephone number of the Society Headquarters is
change between 2005 and 2010 with 250m MODIS data,” addresses key 301-493-0290; the fax number is 301-493-0208; web address is www.asprs.org.
issues of accurately characterizing and quantifying land cover and land PE&RS. PE&RS (ISSN0099-1112) is published monthly by the American
cover change at a continental scale on an operational basis. Society for Photogrammetry and Remote Sensing, 5410 Grosvenor Lane, Suite
210, Bethesda, Maryland 20814-2144. Periodicals postage paid at Bethesda,
The North American Land Change Monitoring System (NALCMS) is Maryland and at additional mailing offices.
specifically needed for applications that require consistent continental land
SUBSCRIPTION. For the 2014 subscription year, ASPRS is offering two
cover information that transcends national boundaries. One example, and options to our PE&RS subscribers -- an e-Subscription and the print edition.
a personal favorite, from the natural resource planning and management E-subscribers can plus-up their subscriptions with printed copies for a small
arena is the fascinating migratory phenomenon of monarch butterflies, a additional charge. Print subscriptions are on a calendar-year basis that runs
from January through December. Electronic subscriptions run for twelve
nearly 2,000 mile journey from the United States and Canada to central months on an anniversary basis. We recommend that customers who choose
Mexican forests. This paper, a culmination of the efforts of the most both e-Subscription and print (e-Subscription + Print) renew on a calendar-year
distinguished remote sensing scientists in three countries and coordinated basis. The new electronic subscription includes access to ten years’ of digital
back issues of PE&RS for online subscribers through the same portal at no
by the Commission for Environmental Cooperation (CEC) based in Montreal, additional charge. Please see the Frequently Asked Questions about our
Canada, presents an excellent example of transboundary collaboration to journal subscriptions.
address North American land change monitoring issues. Pre-processing The rate of the e-Subscription (digital) Site License Only for USA and For-
and classification of remotely sensed data was performed using a scientific eign: $2000; e-Subscription (digital) Site License Only for Canada*: $2100;
yet practical approach. Pre-processing was performed at a central location, Special Offers: e-Subscription (digital) Plus Print for the USA: $2,160;
e-Subscription (digital) Plus Print Canada*: $2,273.25; e-Subscription (dig-
and regionally tuned classification and change approaches were then used ital) Plus Print Outside of the USA: $2,175; Printed-Subscription Only for
to capture specific changes in each country. Having realized the need for USA: $660; Printed-Subscription Only for Canada*: $762; Printed-Sub-
quality pre-processed input data, the Canada Center for Remote Sensing scription Only for Other Foreign: $735. *Note: e-Subscription/Printed-Sub-
scription Only/e-Subscription Plus Print for Canada include 5% of the total
(CCRS) produced monthly composites from MODIS images using an amount for Canada’s Goods and Services Tax (GST #135123065).
improved algorithm. Similarly, each country used different land cover change
POSTMASTER. Send address changes to PE&RS, ASPRS Headquarters, 5410
methods to suit local conditions while retaining the product consistency Grosvenor Lane, Suite 210, Bethesda, Maryland 20814-2144. CDN CPM
at the continental level. This is important because a single continental #(40020812)
algorithm is unlikely to capture the details discerned in this study. Both MEMBERSHIP. Membership is open to any person actively engaged in the prac-
abrupt and gradual changes were monitored. Major outcomes of the paper tice of photogrammetry, photointerpretation, remote sensing and geographic
information systems; or who by means of education or profession is interested
include: improvements to the 2005 base map, country-specific changes
in the application or development of these arts and sciences. Membership is
between this improved map and the updated land cover for the year 2010, for one year, with renewal based on the anniversary date of the month joined.
class-specific changes, site-specific analysis of detected land cover changes, Membership Dues include a 12-month subscription to PE&RS. Subscription
and preliminary results for accuracy assessment. Land cover change is rare. is part of membership benefits and cannot be deducted from annual dues.
Beginning with the January 2014 issue of PE&RS, all members outside of the
The paper revealed that only 1.2% of the North American continent changed USA will receive access to the full digital edition of the journal rather than the
from 2005 to 2010. As expected, rates and types of change vary by country. printed copy. Dues for ASPRS Members outside of the U.S. will now be the
The authors provided a detailed change analysis synopsis in three case study same as for members residing in the U.S. Annual dues for Regular members
(Active Member) is $150; for Student members it is $50 for USA and Canada;
sites for each participating country (Canada, U.S. and Mexico). The research $60 for Other Foreign (E-Journal – No hard copy for all Students); for Associate
also highlights land cover change detection issues and opportunities using Members it is $100 (member must be under the age of 35, see description on
MODIS 250-m resolution satellite data. Change detection issues include application in the back of this Journal). A tax of 5% for Canada’s Goods and
Service Tax (GST #135123065) is applied to all members residing in Canada
spectral differences in input data and small changes that are not discernible
in MODIS data due to its coarse spatial resolution. However, high temporal COPYRIGHT 2014. Copyright by the American Society for Photogrammetry
and Remote Sensing. Reproduction of this issue or any part thereof (except
frequency MODIS data have distinct advantages over other data sources, as short quotations for use in preparing technical and scientific papers) may be
daily image acquisition increases the probability for cloud-free observations made only after obtaining the specific approval of the Managing Editor. The
at various stages of ecosystem changes. Moving forward, the research team Society is not responsible for any statements made or opinions expressed in
technical papers, advertisements, or other portions of this publication. Printed
plans to produce annual land cover and land cover change products on an in the United States of America.
operational basis. Recent advancement in satellite technology, computer
PERMISSION TO PHOTOCOPY. The appearance of the code at the bottom of
technology, free satellite data availability, and integration of science and the first page of an article in this journal indicates the copyright owner’s
engineering will help achieve this goal. The research findings of this paper consent that copies of the article may be made for personal or internal use
and planned research will contribute significantly to improve our scientific or for the personal or internal use of specific clients. This consent is given on
the condition, however, that the copier pay the stated per copy fee of $3.00
understanding of North American land cover and land cover change and its through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers,
implications to our nature and society. Massachusetts 01923, for copying beyond that permitted by Sections 107 or
108 of the U.S. Copyright Law. This consent does not extend to other kinds of
Chandra Giri copying, such as copying for general distribution, for advertising or promotional
U.S. Geological Survey purposes, for creating new collective works, or for resale.
916 October 2014 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
INDUSTRYINSIGHT
This is an open Letter to the Editor—Grids & Datums
invitation to write a
Dear Professor Mugnier,
Letter to the Editor
Often times I am working with historic map sources and in attempting to research
What do you more about them I stumble across the Photogrammetric Engineering and Remote
want to say in Sensing Journal online, which has proven quite resourceful a number of times. I am
250 words or contacting you after reading your May 2008 contribution to the journal -- Grids &
fewer? You pick Datums: Kingdom of Cambodia. I was hoping to get your opinion/recommendation on
the topic, but the correct datum for a 1907-08 map series of French Indochina, at 1:200,000 scale.
keep it on point
I am familiar with the use of Grads on these early 1900s French maps, however this
Let me
particular series does not use them, and shows graticule at 20 minute intervals. There
start by
is also a note beneath the scale bar -- Developpement sinusoidal suivant le meridien
complimenting
101° 40’ which I have assumed refers to the map being in Sinusoidal projection with
the PE&RS editors, namely Russ
101° 40’ central meridian. Based on grid labels I am also confident that this map uses
Congalton, Mike Renslow, and Jie
Paris as prime meridian.
Shan. Our impact factor continues
Do you have any recommendations on a datum the French may have used in this
to climb and is now over 2.0.
region at this time? I appreciate your time, and your contribution to PE&RS journal!
The editors put great effort into
offering authors quality feedback Best Regards,
and support, so good job. Let me
K. Danaher, International Mapping
also compliment the editorial and
the PE&RS production staff and
columnists, namely Rae Kelley, Dear Mr. Danaher,
Matt Austin, Jeannie Congalton,
Since you think they likely used Paris as their prime meridian, my first guess would
Jim Peters, Cliff Mugnier, and
be Belvedere Datum in Vietnam. You can still find the origin point in Hanoi with
Qassim Absullah. In addition to
Google Earth at Mirador Tower at φ = 21º 01′ 57.27″N, λ =105º 50′ 23.33″E (WGS84
the regular columns like Grids and o
Datum/Ersatz Google datum). Original ellipsoid was likely Clarke 1880 (IGN).
Datums and Mapping Matters,
I have been excited to see the Note that at 1:200,000 scale it’s pretty much impossible to identify datum and/or
new columns, such as Member ellipsoid used. It DOES impress the clientele, though.
Interviews, and Behind the Scenes. Best regards,
I like having both peer-reviewed
Clifford J. Mugnier, C.P., C.M.S.
technical papers and increased
Louisiana State University
general interest content together in
the same journal—more for all.
Now let me ask a question. What
do you think of the notion of ASPRS
Attention Students
Special Interest Groups (SIGs)?
Right now, we have technical
divisions, technical committees,
committees of the bylaws, councils,
regions, task forces, and working
groups. Do we have too much
structure encoded into the bylaws
that makes it difficult for us to be n
The ASPRS Scholarships application
agile? What do members think is o
the minimal structure that would deadline for the 2015 awards year is
i
be desirable for ASPRS to function t October 31, 2014.
a
as a society?
c For details, go to:
By the way, this letter to the
u www.asprs.org/Asprs-Awards-And-
editor is exactly 250 words, so as
d Scholarships.html
you can see, writers need to write
e
on point.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING October 2014 917
D n A
etection of orth mericAn
L c c
AnD over hAnge
B 2005 2010
etween AnD
250 moDiS D
with m AtA
By René R Colditz, Darren Pouliot, Ricardo M. Llamas, Collin Homer, Rasim Latifovic, Rainer
A. Ressl, Carmen Meneses Tovar, Arturo Victoria Hernández, and Karen Richardson
i
ntroDuction
L
Land cover and land cover change data are needed for with 19 classes was derived from 250m Moderate Resolution
climate change, biodiversity and ecosystem studies. Imaging Spectroradiometer (MODIS) image composites and
Consistent multi-temporal mapping and monitoring published in 2010 (Latifovic et al. 2012). Generating maps
of land cover change over large regions remains to at the continental scale rather than using global products
be a challenging task in remote sensing science. The ensures the best-available, consistently-produced map for this
North American Land Change Monitoring System geographic extent (Latifovic et al. 2012, Colditz et al. 2012).
(NALCMS) is a collaborative effort among governmental Since then, NALCMS has focused on change detection with
institutions of Canada, the United States and Mexico the long-term goal to produce a consistent annual time series
to monitor the state of land cover across the North of land cover data. NALCMS land cover products are designed
American continent. The initiative is facilitated by the to meet the needs of communities that require consistent
Commission for Environmental Cooperation (CEC), an land cover change data at a continental scale. These include
international organization created by the ministries of applications in climate change, weather service, hydrology,
environment of Canada, Mexico, and the United States biodiversity and carbon sequestration. This paper presents
to promote environmental collaboration among the three the NALCMS approach for change detection and results of
countries. Participating institutions in NALCMS are the land cover change analysis between the years 2005 and
Natural Resources Canada / Canada Centre for Remote 2010.
Sensing (NRCan / CCRS), the United States Geological
Survey (USGS), and for Mexico the National Institute Consistent multi-temporal mapping and
for Statistics and Geography (INEGI), the National monitoring of land cover change over large
Commission for the Knowledge and Use of Biodiversity
regions remains to be a challenging task in
(CONABIO), and the National Forestry Commission
remote sensing science.
(CONAFOR). A 2005 land cover map of North America
918 October 2014 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
D m Additional post-processing steps involved evidence-based
AtA AnD ethoDS
spatial and contextual reasoning for Canada, localized GIS
Change detection and labeling were both based on monthly
modeling in the U.S., and a transition matrix to only allow
continental composites of MODIS reflectance observations
logical changes in Mexico.
acquired for 2005 and 2010. The composites were produced by
the CCRS following methodologies described in Khlopenkov and
Identifying land cover change required a
Trishchenko (2008), Luo et al. (2008), and Latifovic et al. (2012).
two-step protocol: (1) detecting areas of
Identifying land cover change required a two-step protocol: (1)
detecting areas of potential change and (2) assigning a land potential change and (2) assigning a land
cover class to those areas, similar to that described in Latifovic
cover class to those areas.
and Pouliot (2005), Xian et al. (2009), Fry et al. (2011), Pouliot
et al. (2012), and Pouliot et al. (2014). Unlike other approaches
r
eSuLtS
such as the global MODIS land cover maps (MCD12Q1; Friedl
et al. 2010), this two-step approach assures high consistency Results are presented in the following four sections: (1)
of mapping in a land cover time series, because areas where improvements to the 2005 base map, (2) country-specific
change has not occurred remain unaffected and change areas changes between this improved map and the updated
can be directly compared across time. land cover for year 2010, including preliminary results for
Prior to change detection, the initial North America land accuracy assessment, (3) class-specific changes, and (4) site-
cover map of 2005 (LC2005) was modified, and the improved specific analysis of detected land cover changes. The results
version (LC2005V2) was used as baseline for generating the presented here were derived for classifications at level 2 (19
2010 land cover product. This was necessary because the classes) and a minimum mapping unit of 25ha. The state of
updating procedure required an accurate baseline as only Hawaii was excluded from this analysis.
areas of potential change will be altered and mislabeled
Difference Between the Original and Improved Land
patches or false detection would have negatively affected
Cover Maps of 2005
change results in their spatial context.
The actual implementation of the updating procedure varied A total area of 1,077,701km2 (5.06% of the North American
among the countries because of existing prior approaches continent) was modified in the improved base line map
and additional custom needs of each participating country. (LC2005V2). For Canada, further quality assessment and
Country-specific procedures did not notably affect the overall input from users suggested modifications in some areas (in
map consistency. For the first step of change detection, the total 638,963km2, 6.45% of the total area of Canada). The
binary mask of change and no change for Canadian forests major changes included reduction of treed wetland along
was derived using a regression tree procedure for abrupt the western and eastern coasts, correction of water/shadow
disturbances (Guindon et al. 2011) and a multifaceted trend- confusion in mountain areas, enhanced shrub classification,
based approach for detection of more gradual and subtle and some local improvements to wetland mapping.
changes (Pouliot et al. 2014). The United States applied a Improvements in the United States entirely focused on the
complex change vector approach mainly based on the NDVI state of Alaska, which involved a complete reprocessing due
and red and near infrared bands (Jin et al. 2013). For Mexico, to a spatial mis-registration of the initial 2005 land cover map
a data-driven process that uses difference images of all for this state (429,916km2, 4.55% of the total area of the U.S.),
bands, the NDVI, and edge detection filters from all months while the land cover map of the lower 48 states was deemed
was developed (Colditz et al. 2014). sufficient for map updating. In Mexico, improvements focused
For the second step of change labeling, each country on two land cover classes (8,821km2, 0.45% of the total area
also conducted its own specific classification approach yet of Mexico): urban and built-up areas were underestimated or
followed the same general procedure: training data were omitted due to a low spectral contrast to sparse vegetation
sampled from surrounding unchanged areas and the spectral in semi-arid northern Mexico and class Water, which was
data set of 2010 served as features to generate a decision tree generally underestimated in the original map.
that was subsequently applied to the potential change areas.
PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING October 2014 919
Country-specific Change and its Accuracy southern British Columbia (Bleiker et al. 2011). In the Unit-
T
Between 2005 and 2010 ed States, fire is also the greatest source of land cover change
The total area of detected land cover changes between in this database, with the vast majority of areas occurring
2005 and 2010 was 259,285km2 which corresponds to in the western U.S. and Alaska. Other types of land cover
1.22% of the North American continent (Table 1). Al- change do occur, but are not as well represented as higher
most 80% of all detected changes occurred in Canada, resolution databases have indicated (Fry et al., 2011). Expan-
another 20% in the United States, and only 1.1% in sion of urban area and variations in water level of artificial
Mexico. In Canada, forest fires are the major source of lakes are the most frequently detected changes in Mexico,
change (NFD 2013) and are a natural process essential while forest loss due to fire or harvest was underestimated.
to ecosystem functions (Weber and Stocks 1998). Most Smaller but economically important detected changes were
fires in Canada are stand replacing and thus several expanding mining sites in northern Mexico and changes in
stages of change are evident throughout a forest succes- rain-fed agricultural zones in northern and central Mexico.
sion pathway. Forest The main limitation for change detection, besides
harvesting is anoth- Table 1. Area of change in square a significant spectral difference in image data and an
kilometers and percent between the
er agent of change land cover map of 2005 and 2010. appropriate algorithm, was the size of the change object on
as well as insect-in- the ground (Pouliot et al. 2009). In this respect this study
Area km2 %
duced mortality. The is limited by the moderate spatial resolution of 250m. While
North America 259,285 1.22
mountain pine beetle this resolution was sufficient to detect large-area changes in
Canada 204,049 2.06
infestation is respon- Canada (70-80% overall accuracy, Pouliot et al. 2014) and
United States 52,294 0.55
sible for much of the the western United States, many changes in the complex
Mexico 2,942 0.15
change observed in and small-patch landscapes of the eastern United States
Table 2. Summary statistics from land cover maps and change matrix with area gain, loss, and balance.
LC2005V2 LC2010 Gain Loss Balance
km2 % km2 % km2 % km2 % km2 %
Temperate or sub-polar needleleaf forest 3,209,808 15.06 3,133,399 14.70 14,418 5.56 90,828 35.03 -76,410 -29.47
Sub-polar taiga needleleaf forest 522,951 2.45 518,359 2.43 9,637 3.72 14,230 5.49 -4,593 -1.77
Tropical or sub-tropical broadleaf evergreen forest 155,347 0.73 155,104 0.73 16 0.01 258 0.10 -242 -0.09
Tropical or sub-tropical broadleaf deciduous forest 162,700 0.76 162,627 0.76 23 0.01 96 0.04 -73 -0.03
Temperate or sub-polar broadleaf deciduous forest 1,617,791 7.59 1,641,647 7.70 34,490 13.30 10,634 4.10 23,856 9.20
Mixed forest 1,386,319 6.50 1,385,318 6.50 31,792 12.26 32,793 12.65 -1,001 -0.39
Tropical or sub-tropical shrubland 960,715 4.51 960,577 4.51 1,208 0.47 1,346 0.52 -138 -0.05
Temperate or sub-polar shrubland 2,615,944 12.27 2,640,422 12.39 69,784 26.91 45,306 17.47 24,478 9.44
Tropical or sub-tropical grassland 33,523 0.16 33,619 0.16 386 0.15 290 0.11 96 0.04
Temperate or sub-polar grassland 1,634,507 7.67 1,673,703 7.85 75,569 29.14 36,373 14.03 39,196 15.12
Sub-polar or polar shrubland-lichen-moss 433,470 2.03 432,423 2.03 167 0.06 1,215 0.47 -1,048 -0.40
Sub-polar or polar grassland-lichen-moss 1,239,877 5.82 1,241,231 5.82 1,926 0.74 572 0.22 1,354 0.52
Sub-polar or polar barren-lichen-moss 615,865 2.89 615,861 2.89 11 0.00 15 0.01 -4 0.00
Wetland 750,883 3.52 744,589 3.49 3,267 1.26 9,561 3.69 -6,294 -2.43
Cropland 2,947,602 13.83 2,944,167 13.81 7,901 3.05 11,337 4.37 -3,436 -1.32
Barren land 1,191,144 5.59 1,191,846 5.59 2,181 0.84 1,480 0.57 701 0.27
Urban and built-up 238,830 1.12 239,590 1.12 1,565 0.60 805 0.31 760 0.29
Water 1,292,067 6.06 1,294,890 6.07 4,924 1.90 2,101 0.81 2,823 1.09
Snow and ice 309,433 1.45 309,405 1.45 19 0.01 46 0.02 -27 -0.01
Note: Areas in square kilometers are rounded. Percentages of classes in the land cover maps are relative to the entire study area (21,318,777 km2) but
for gain, loss and balance they are relative to the total change area (259,285 km2).
920 October 2014 PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING