<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Mike Wulder on Medium]]></title>
        <description><![CDATA[Stories by Mike Wulder on Medium]]></description>
        <link>https://medium.com/@mikewulder?source=rss-39acee693c3a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*ldYRXygZHTKdEgjI.jpg</url>
            <title>Stories by Mike Wulder on Medium</title>
            <link>https://medium.com/@mikewulder?source=rss-39acee693c3a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 20:24:47 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@mikewulder/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Landsat data record for Canada: An update]]></title>
            <link>https://medium.com/@mikewulder/landsat-data-record-for-canada-an-update-38b176f49a4f?source=rss-39acee693c3a------2</link>
            <guid isPermaLink="false">https://medium.com/p/38b176f49a4f</guid>
            <category><![CDATA[monitoring]]></category>
            <category><![CDATA[forest]]></category>
            <category><![CDATA[satellite-technology]]></category>
            <category><![CDATA[landsat]]></category>
            <category><![CDATA[canada]]></category>
            <dc:creator><![CDATA[Mike Wulder]]></dc:creator>
            <pubDate>Wed, 07 Feb 2018 00:14:10 GMT</pubDate>
            <atom:updated>2018-02-07T00:14:10.332Z</atom:updated>
            <content:encoded><![CDATA[<p>In 2014 we published an accounting of Landsat images available for Canada (White and Wulder 2014) and did some analyses to better understand the spatial and temporal distribution of these images. The material in White and Wulder (2014) represented the period from 1972 to 2012. For science and applications we needed information regarding where the images are located and when they were acquired (both within year and between years) (White et al. 2014). Knowledge of the depth and breadth of the USGS image holdings in a readily accessible, free and open access form encouraged our scientific, methodological, and applications work (in collaboration with the <a href="https://medium.com/u/34fa161719a5">Integrated RS Studio</a> and with acceleration funding from the Canadian Space Agency) to recreate the history of Canada’s forests (Hermosilla at al. 2016) and make the outcomes of this research publicly available (<a href="https://opendata.nfis.org/mapserver/nfis-change_eng.html">data</a>, <a href="https://www.researchgate.net/publication/318380375_Integrating_the_past_present_future_of_Landsat_Continuity_of_science_applications_monitoring_and_reporting">science</a>).</p><p>On the 5th anniversary of the February 11, 2013 launch of Landsat-8 we thought it would be good to reflect on the current status of the Landsat data record. Since 2013, over 1.1 million scenes have been acquired globally by Landsat-8 (<a href="https://landsat.usgs.gov/download-entire-collection-metadata">USGS Landsat metadata</a>) adding to the already expansive Landsat archive, which started almost 45 years ago (<a href="https://landsat.usgs.gov/sites/default/files/documents/1-s2.0-S0034425715302194-main.pdf">Wulder et al. 2016</a>). The result is the addition of ~220,000 OLI images globally per year. Since the 2013 Landsat-8 launch, over 105,700 OLI images have been acquired for Canada. This translates into an average of approximately 22,500 Landsat-8 OLI images per year, or more than 60 per day. This is a good news story for Canada, but it gets even better: Landsat-7 continues to operate and add images to the global archive.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/626/1*SXSrkqNCY8eh_Dy9mtsZ8g.png" /><figcaption>Figure 1. Landsat-7 and -8 images representing Canada added to the USGS Landsat archive for the years including 2013 to 2017.</figcaption></figure><p>With both Landsat-7 ETM+ and Landsat-8 in operation approximately 1200 images are collected globally per day: &gt;430 ETM+ and &gt;740 OLI. The Landsat long-term <a href="https://landsat.usgs.gov/acquisition">acquisition plan</a> (LTAP) aids in guiding collections globally and uses knowledge of cloud cover and intelligent, spatially informed, scheduling to distribute image collects globally. Based upon a large on-board data recorder and high-speed downlink for imagery (Irons et al. 2012), Landsat-8 can operate in near “always on” mode for earth’s terrestrial land mass. This global yield when considered over Canada, has resulted in &gt;38,500 images per year total, broken down as &gt; 16,000 ETM+ and &gt;22,500 OLI. Following satellite check-in and confirmation of stable operations, testing of of critical mission (ground) operations systems was undertaken to determine the capacity to increase Landsat-8 collections. An ability to collect, transmit, and process image numbers in excess of specification was confirmed, allowing for additional collections to be made, as evident below in Figure 1. Nearly one quarter of all the images in the USGS archive for Canada were acquired over the 2013 to 2017 (10% ETM+ and 13.4% OLI).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/901/1*BsXZmBElG4u03B7G6Nc5cQ.png" /><figcaption>Figure 2. Total number of archived Landsat images acquired for Canada, by year and sensor.</figcaption></figure><p>Given the best information available at the time (related to orbital status and fuel level, <a href="https://landsat.gsfc.nasa.gov/contingency-plans-for-landsat-5-7/">link</a>), we reported in White and Wulder (2014) an expectation of end of service of Landsat-7 to occur in 2017. Based upon revisions to fuel use scenarios and decommissioning plans, to the best of current understanding, the end-of-life expectation for Landsat-7 is now projected to be Autumn of 2020 (<a href="https://www.researchgate.net/publication/317607241_Landsat_Science_Team_2017_Winter_Meeting_Summary">link</a>). The launch of Landsat-9 is tentatively scheduled for December 2020. Note, Landsat-8 has a 5 year design life and 10 years of fuel (Irons et al. 2012).</p><p>Landsat is of unparalleled global importance. Acquisitions are only one aspect of a satellite program. The systematic collection, downlink, preparation, and distribution of Landsat data all combine to support science and application developments. Global information aspirations can be met with Landsat as well as national and regional science and application needs. The increasingly long baseline provided by Landsat provides Canada with an otherwise unavailable resource, allowing for the generation of information products to describe and inform on a complex and dynamic country.</p><h3>Resources</h3><p>Hermosilla, T., Wulder, M.A., White, J.C., Coops, N.C., Hobart, G.W. and Campbell, L.B., 2016. <a href="http://www.tandfonline.com/doi/full/10.1080/17538947.2016.1187673">Mass data processing of time series Landsat imagery: pixels to data products for forest monitoring</a>. <em>International Journal of Digital Earth</em>, <em>9</em>(11), pp.1035–1054. #OpenAccess: <a href="http://bit.ly/29XRECd">http://bit.ly/29XRECd</a></p><p>Irons, J.R., Dwyer, J.L. and Barsi, J.A., 2012. <a href="https://www.sciencedirect.com/science/article/pii/S0034425712000363">The next Landsat satellite: The Landsat data continuity mission</a>. <em>Remote Sensing of Environment</em>, <em>122</em>, pp.11–21.</p><p>Roy, D.P., Wulder, M.A., Loveland, T.R., Woodcock, C.E., Allen, R.G., Anderson, M.C., Helder, D., Irons, J.R., Johnson, D.M., Kennedy, R. and Scambos, T.A., et al. 2014. <a href="https://www.sciencedirect.com/science/article/pii/S003442571400042X">Landsat-8: Science and product vision for terrestrial global change research</a>. <em>Remote sensing of Environment</em>, <em>145</em>, pp.154–172. #OpenAccess: <a href="http://bit.ly/2El8i1v">http://bit.ly/2El8i1v</a></p><p>White, J.C. and Wulder, M.A., 2014. <a href="http://www.tandfonline.com/doi/abs/10.5589/m13-053">The Landsat observation record of Canada: 1972–2012</a>. <em>Canadian Journal of Remote Sensing</em>, <em>39</em>(6), pp.455–467. #OpenAccess: <a href="http://bit.ly/2BHqTnV">http://bit.ly/2BHqTnV</a></p><p>White, J.C., Wulder, M.A., Hobart, G.W., Luther, J.E., Hermosilla, T., Griffiths, P., Coops, N.C., Hall, R.J., Hostert, P., Dyk, A. and Guindon, L., 2014. <a href="http://www.tandfonline.com/doi/full/10.1080/07038992.2014.945827">Pixel-based image compositing for large-area dense time series applications and science</a>. <em>Canadian Journal of Remote Sensing</em>, <em>40</em>(3), pp.192–212. #OpenAccess: <a href="http://bit.ly/2AP6s3E">http://bit.ly/2AP6s3E</a></p><p>Wulder, M.A., White, J.C., Loveland, T.R., Woodcock, C.E., Belward, A.S., Cohen, W.B., Fosnight, E.A., Shaw, J., Masek, J.G. and Roy, D.P., 2016. <a href="https://landsat.usgs.gov/sites/default/files/documents/1-s2.0-S0034425715302194-main.pdf">The global Landsat archive: Status, consolidation, and direction</a>. <em>Remote Sensing of Environment</em>, <em>185</em>, pp.271–283. #OpenAccess: <a href="http://bit.ly/2xoLS8W">http://bit.ly/2xoLS8W</a></p><h4>Links</h4><p>High Resolution Forest Change for Canada (data)<br><a href="https://opendata.nfis.org/mapserver/nfis-change_eng.html">https://opendata.nfis.org/mapserver/nfis-change_eng.html</a></p><p>Summary of Landsat supported forest monitoring science and applications for Canada<br><a href="http://bit.ly/2t4ck9W">http://bit.ly/2t4ck9W</a></p><p>USGS Landsat image metadata<br><a href="https://landsat.usgs.gov/download-entire-collection-metadata">https://landsat.usgs.gov/download-entire-collection-metadata</a></p><p>NASA info on Landsat-8<br><a href="https://landsat.gsfc.nasa.gov/landsat-8/">https://landsat.gsfc.nasa.gov/landsat-8/</a></p><p>NASA Landsat Program (Twitter): <a href="https://twitter.com/NASA_Landsat">@NASA_Landsat</a></p><p>USGS Landsat Program (Twitter): <a href="https://twitter.com/USGSLandsat">@USGSLandsat</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=38b176f49a4f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Canada’s forests return following disturbance]]></title>
            <link>https://medium.com/@mikewulder/canadas-forests-return-following-disturbance-fff980ada16c?source=rss-39acee693c3a------2</link>
            <guid isPermaLink="false">https://medium.com/p/fff980ada16c</guid>
            <category><![CDATA[forest-ecology]]></category>
            <category><![CDATA[satellite-technology]]></category>
            <category><![CDATA[climate-change]]></category>
            <category><![CDATA[forestry]]></category>
            <dc:creator><![CDATA[Mike Wulder]]></dc:creator>
            <pubDate>Tue, 11 Apr 2017 00:03:58 GMT</pubDate>
            <atom:updated>2017-04-11T13:38:04.102Z</atom:updated>
            <content:encoded><![CDATA[<p>In a recently published paper, <a href="https://www.researchgate.net/profile/Joanne_White2">Joanne White</a> (Research Scientist, Canadian Forest Service) and co-authors communicate their findings characterizing 25 years of forest disturbance and recovery for Canada using satellite data. Using open data from the Landsat series of satellites and high performance computing opportunities, White and co-authors use tens-of-thousands of images representing billions of pixels, to map forest wildfire and harvesting on an annual basis and to then characterize the return of forest vegetation following the disturbance. Forest disturbance is captured and mapped to type (such as wildfire or harvest). Using the measurements from the post-disturbance satellite data, it is found that given time forests return following disturbance. As Joanne indicates,</p><blockquote>“The capacity to distinguish recovery trends by disturbance type provides insights on forest dynamics that are both spatially detailed and national in scope.”</blockquote><p><strong>What are the major findings of this research?</strong></p><ul><li>57.5 Mha or 10.75% of Canada’s net forested ecosystems (exclusive of water) were disturbed by either wildfire or harvest, representing an annual rate of disturbance of approximately 0.43% per year.</li><li>Wildfire accounted for 2.5 times more area disturbed than harvest. On average, wildfire impacted 1.5 Mha annually, compared to 0.65 Mha impacted by harvesting. Moreover, the amount of area impacted by wildfire was much more variable annually (σ = 1.5 Mha) when compared to harvest (σ= 0.1 Mha).</li><li>Regionally, boreal ecozones had the greatest absolute area of disturbance from both wildfire (Boreal Shield West) and harvest (Boreal Shield East).</li><li>Spectral recovery metrics indicate that vegetation returns following disturbance.</li><li>Areas impacted by harvest exhibited more rapid spectral recovery rates.</li><li>78.6% of areas impacted by harvest having a mean recovery time ≤ 10 years, compared to 35.5% of areas impacted by wildfire</li><li>Overall, less than 1% of the areas disturbed by wildfire and harvest were identified as non-recovering by all three spectral measures of recovery used in our analysis.</li></ul><p><strong><em>Citation and open access:</em></strong></p><p>White, J.C., Wulder, M.A., Hermosilla, T., Coops, N.C., Hobart, G.W. 2017. A nationwide annual characterization of 25 years of forest disturbance and recovery for Canada using Landsat time series. Remote Sensing of Environment, 194: 303–321. DOI: <a href="http://doi.org/10.1016/j.rse.2017.03.035">10.1016/j.rse.2017.03.035</a></p><p>Get it here: <a href="https://authors.elsevier.com/sd/article/S0034425717301360">https://authors.elsevier.com/sd/article/S0034425717301360</a></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/913/1*UusMBWF5HCZAPBZMQgfiAw.gif" /><figcaption>Best Available Pixel (BAP) proxy image composite of Canada in 2010 (Landsat TM/ETM+bands 5,4,3) superimposed with the boundaries of the forested ecozones analyzed in this study (A). Area disturbed by wildfire and harvest (1985–2010), as identified using the Composite to Change (C2C) approach (B).</figcaption></figure><p>Vegetation recovery following disturbance is a process that varies greatly in both time and space. A 5-year window, as measured by our absolute or relative short-term spectral recovery metrics provides an initial assessment of forest conditions post disturbance. To evaluate recovery over a longer temporal window, we calculated how many years it took for a pixel to reach 80% of its pre-disturbance Normalized Burn Ratio (NBR) value (Y2R). The pre-disturbance NBR value is determined using the average NBR value for the two years prior to disturbance.</p><p><strong>Every paper has a key figure that really brings home the key findings. In this paper, it is this figure showing the years to recovery following disturbance (by type), with the recovery trends also stratified by epochs.</strong></p><blockquote><strong>So, Joanne, what does this figure show?</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9jMU1KETj-9J7ZmsNSZTkQ.jpeg" /><figcaption>National summary of Years to Recovery (Y2R) values for wildfire and harvest. The cumulative proportion of area disturbed by Y2R is plotted for all disturbances (1985–2005), for disturbances that occurred prior to 1990, and for disturbances that occurred between 1990 and 2000</figcaption></figure><blockquote>“In the context of our definition of recovery, and given the emphasis that we place on characterizing relative rates of recovery, meeting or exceeding 80% of the pre-disturbance NBR value relates a positive trend in the return of vegetation, but does not necessarily indicate a return to the same forest conditions that existed at a site prior to disturbance.”</blockquote><p>Further, Joanne offers, “Our results indicated that there was a difference in long-term recovery between areas disturbed by wildfire and harvest. We found that approximately, 68.4% of wildfire areas and 92.5% of harvest areas attained the 80% benchmark by the end of the time series in 2010. It should be noted that this result includes all disturbances that occurred between 1985 and 2005, so for some disturbances, there is only a 5-year period post-disturbance available for analysis. Therefore, we specifically considered only those disturbances in our time series that have had the longest time period of recovery (that is, those disturbances that occurred prior to 1990). When we focused on this subset of disturbances, we found that 86.3% of wildfire areas, and 98.4% of harvest areas attained an NBR value that was 80% of their respective pre-disturbance NBR.”</p><p><strong>What are the unique contributions of this research?</strong></p><p>• Remote sensing-informed assessment of 25 years of wildfire and harvest in Canada’s forested ecosystem (not done previously from Landsat time series at 30 m spatial resolution);</p><p>• National, spatially-explicit record of harvesting (no such record exists in Canada);</p><p>• Characterize national trends in recovery following wildfire and harvest, which has previously only been achieved on a sample-basis or by using AVHRR data;</p><p>• Characterize trends in Canada’s forest recovery separately by disturbance type (wildfire and harvest), which has likewise not been previously undertaken.</p><p><strong>Background:</strong></p><p>The objective of this <a href="https://twitter.com/NRCan">@NRCan</a> research, accelerated by support from the Canadian Space Agency (<a href="https://twitter.com/csa_asc">@CSA_ASC</a>), and through collaboration with the University of British Columbia (Prof Nicholas Coops, <a href="https://twitter.com/IRSS_UBC">@IRSS_UBC</a> and Dr. Txomin Hermosilla, <a href="https://twitter.com/txominhermos">@txominhermos</a>) was to characterize national trends in stand replacing forest disturbance caused by wildfire and harvest, and subsequent recovery, for the period 1985–2010 for Canada’s forested ecosystems (~650 Mha), using information derived from Landsat time series data. These results represent a major information outcome from the Composite-2-Change (C2C) approach of generating best-available-pixel (BAP) composites from Landsat data, and identifying and attributing forest change on an annual basis (as presented in Hermosilla et al. 2016). This publication is important because it is national in scope, and represents the first wall-to-wall characterization of wildfire and harvest and subsequent vegetation recovery in Canada at a spatial resolution commensurate with human impacts. The information outcomes represent 25 years of change in Canada’s forests, derived from a single, consistent spatially-explicit data source, derived in a fully automated manner. This demonstrated capacity to characterize forests at a resolution that captures human impacts is key to establishing a baseline for current and future climate change monitoring. Moreover, characterizing both the depletion (disturbance) and accrual (regrowth) of vegetation provides a more holistic understanding forest change in the context of long-term forest monitoring and carbon accounting. The high performance computing environment of <a href="https://twitter.com/WestGrid">@WestGrid</a>, a part of <a href="https://twitter.com/ComputeCanada">@ComputeCanada</a>, made this research and collaboration possible. Further, the free and open access data policy of the United States Geological Survey (<a href="https://twitter.com/USGSLandsat">@USGSLandsat</a>) provided the data (from the <a href="https://twitter.com/NASA_Landsat">@NASA_Landsat</a> series of Landsat satellites) in a robust, analysis ready form, promoting our scientific analyses.</p><p><strong>How do the C2C outputs compare to other data sources?</strong></p><p>In the paper, remotely sensed outputs are compared to statistics for wildfire fire and harvest from the <a href="https://t.co/Zi7BRi0UCr">National Forestry Database</a> and <a href="http://cwfis.cfs.nrcan.gc.ca/ha/nfdb">Canadian National Fire Database</a>. Trends are similar among data sources, although absolute estimates of area differ. In the case of harvests, much of the earlier data used to report national trends in harvesting over time were aspatial. Methods for recording harvest information have evolved over time as GIS technology has become ubiquitous in natural resource management agencies. Despite this, there is currently no national, spatially-explicit record of forest harvesting in Canada. In the case of wildfire, the mapped outcomes are expected to differ, based on known differences in the methods applied and data used, especially as related to differences in source data spatial resolution, definitions of minimum mapable units, and treatment of unburned islands and waterbodies within wildfire boundaries. We also acknowledge that the C2C outputs have a latency issue in detecting disturbance as a function of the image compositing window used (i.e. August 1 ± 30 days); however a national independent accuracy assessment indicated that 97.7% of changes were labelled to within ± 1 year (Hermosilla et al. 2016).</p><p><strong>What about other disturbance types (e.g. insects)?</strong></p><p>While the focus of this analysis was on stand replacing disturbance, other forms of disturbance (e.g. insects, water stress) can also play an important role in the dynamics of forested ecosystems and can have significant impacts on applications such as carbon accounting. While some insect disturbances manifest in ways that are relatively abrupt and readily detectable (e.g. bark beetles) others are more ephemeral (e.g. defoliators) and although separating the two has been demonstrated, it it is challenging to do so in a national context, where there are many potential and region-specific (and often overlapping) insects and pathogens present. The capacity to use Landsat time series analysis to characterize these various non-stand replacing disturbances is emerging; however, spectral measures of recovery from these non-stand replacing changes have yet to be explored and will require further consideration and research.</p><p><strong>How do these satellite measures relate to ground (plot) measurements?</strong></p><p>Information from remotely sensed data can provide a useful framework for assessing relative rates and changes in spectral recovery that enables the national assessment of trends. These trends can be related to available ground observations, and can form the basis for additional sampling and investigations to further relate spectral measures of recovery to ecological and silvicultural understanding of the recovery process. Such national assessments are not possible using existing ground plot measurements alone, as demonstrated in Bartels et al. (2016).</p><p><strong>Key takeaways</strong></p><p>By leveraging recent advances in image compositing capability and the holdings of the Landsat archive, we generated a synoptic, consistent national baseline of stand-replacing forest disturbance and recovery and characterized important regional variations in the disturbance and recovery trends observed, as well as variations related to the disturbance type (i.e. wildfire, harvest). These baseline data provide unprecedented reference information against which present and future trends in disturbance and recovery can be assessed. Efforts are underway to share the data in an open access form. To view the change data online, see: <a href="http://forests.foundryspatial.com/">http://forests.foundryspatial.com/</a>.</p><p><strong>NRCan Information Links:</strong></p><p><a href="http://www.nrcan.gc.ca/forests/fire-insects-disturbances/tracking-change/17825">Tracking forest change and recovery in Canada</a></p><p><a href="http://www.nrcan.gc.ca/forests/report/disturbance/16502">How disturbances shape Canada’s forests</a></p><p><strong>References (peer review):</strong></p><blockquote><em>Multi-decade high-resolution forest disturbance and recovery for Canada:</em></blockquote><blockquote>White, J.C., Wulder, M.A., Hermosilla, T., Coops, N.C., Hobart, G.W. 2017. A nationwide annual characterization of 25 years of forest disturbance and recovery for Canada using Landsat time series. Remote Sensing of Environment, 194: 303–321. DOI: <a href="http://doi.org/10.1016/j.rse.2017.03.035">10.1016/j.rse.2017.03.035</a></blockquote><blockquote>Get it here, #OpenAccess: h<a href="https://authors.elsevier.com/sd/article/S0034425717301360">ttps://authors.elsevier.com/sd/article/S0034425717301360</a></blockquote><p><em>Data used:</em></p><p>White, J. C., &amp; Wulder, M. A. (2014). The Landsat observation record of Canada: 1972–2012. <em>Canadian Journal of Remote Sensing</em>, <em>39</em>(6), 455–467. <a href="https://www.researchgate.net/publication/260163311_The_Landsat_observation_record_of_Canada_1972-2012">Link</a></p><p><em>Project rationale and information on image compositing:</em></p><p>White, J. C., Wulder, M. A., Hobart, G. W., Luther, J. E., Hermosilla, T., Griffiths, P., … &amp; Guindon, L. (2014). Pixel-based image compositing for large-area dense time series applications and science. <em>Canadian Journal of Remote Sensing</em>, <em>40</em>(3), 192–212. <a href="https://www.researchgate.net/publication/267509030_Pixel-Based_Image_Compositing_for_Large-Area_Dense_Time_Series_Applications_and_Science">Link</a></p><p><em>Generation of gap-free surface reflectance composites:</em></p><p>Hermosilla, T., Wulder, M. A., White, J. C., Coops, N. C., &amp; Hobart, G. W. (2015a). An integrated Landsat time series protocol for change detection and generation of annual gap-free surface reflectance composites. <em>Remote Sensing of Environment</em>, <em>158</em>, 220–234. <a href="https://www.researchgate.net/publication/269113059_An_integrated_Landsat_time_series_protocol_for_change_detection_and_generation_of_annual_gap-free_surface_reflectance_composites">Link</a></p><p><em>Labeling of change type using spatial, spectral, temporal, and contextual information:</em></p><p>Hermosilla, T., Wulder, M. A., White, J. C., Coops, N. C., &amp; Hobart, G. W. (2015b). Regional detection, characterization, and attribution of annual forest change from 1984 to 2012 using Landsat-derived time-series metrics. <em>Remote Sensing of Environment</em>, <em>170</em>, 121–132. <a href="https://www.researchgate.net/publication/282327786_Regional_detection_characterization_and_attribution_of_annual_forest_change_from_1984_to_2012_using_Landsat-derived_time-series_metrics">Link</a></p><p><em>Overall methods and accuracy assessment:</em></p><p>Hermosilla, T., Wulder, M. A., White, J. C., Coops, N. C., Hobart, G. W., &amp; Campbell, L. B. (2016). Mass data processing of time series Landsat imagery: pixels to data products for forest monitoring. <em>International Journal of Digital Earth</em>, <em>9</em>(11), 1035–1054. <a href="http://www.tandfonline.com/doi/full/10.1080/17538947.2016.1187673">Link</a></p><p><em>Description of plot based insights on forest recovery in Canada:</em></p><p>Bartels, S. F., Chen, H. Y., Wulder, M. A., &amp; White, J. C. (2016). Trends in post-disturbance recovery rates of Canada’s forests following wildfire and harvest. <em>Forest Ecology and Management</em>, <em>361</em>, 194–207. <a href="https://www.researchgate.net/publication/284358345_Trends_in_post-disturbance_recovery_rates_of_Canada%27s_forests_following_wildfire_and_harvest">Link</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fff980ada16c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Communication of multi-year change]]></title>
            <link>https://medium.com/@mikewulder/communication-of-multi-year-change-32a20e95f76?source=rss-39acee693c3a------2</link>
            <guid isPermaLink="false">https://medium.com/p/32a20e95f76</guid>
            <category><![CDATA[scicomm]]></category>
            <category><![CDATA[forest]]></category>
            <category><![CDATA[gis]]></category>
            <category><![CDATA[remote-sensing]]></category>
            <category><![CDATA[satellite-technology]]></category>
            <dc:creator><![CDATA[Mike Wulder]]></dc:creator>
            <pubDate>Thu, 06 Apr 2017 21:32:00 GMT</pubDate>
            <atom:updated>2017-04-07T03:30:29.039Z</atom:updated>
            <content:encoded><![CDATA[<blockquote>Problem: How to best communicate land cover change over time for multiple years? Any ideas? We need help!</blockquote><blockquote>We have a multi-decade annual time series of land cover maps (a land cover data cube, if you will) and we are interested in insights on how to distill the information present in meaningful, informative, ways.</blockquote><p>We are currently working on recreating the history of Canada’s forests using Landsat data. As outlined in <a href="https://www.researchgate.net/publication/267509030_Pixel-Based_Image_Compositing_for_Large-Area_Dense_Time_Series_Applications_and_Science">White et al. (2014)</a>, information on forest dynamics is required to inform science and reporting activities as well as to address information needs of the National Forest Inventory and Carbon Accounting activities. Previously we have communicated our approach for creating a Canada-wide annual time-series stack of surface reflectance data to compute change over time (<a href="https://www.researchgate.net/publication/269113059_An_integrated_Landsat_time_series_protocol_for_change_detection_and_generation_of_annual_gap-free_surface_reflectance_composites">Hermosilla et al. 2015</a>a) and label these changes to type, such as wildfire or harvest (<a href="https://www.researchgate.net/publication/282327786_Regional_detection_characterization_and_attribution_of_annual_forest_change_from_1984_to_2012_using_Landsat-derived_time-series_metrics">Hermosilla et al. 2015</a>b). The end-to-end process and accuracy assessment of nation-wide outcomes for Canada can be found in <a href="http://www.tandfonline.com/doi/full/10.1080/17538947.2016.1187673">Hermosilla et al. (2016)</a>. The outcomes of this research can be viewed and manipulated at the following site: <a href="http://forests.foundryspatial.com/">http://forests.foundryspatial.com/</a>.</p><p>Having in hand the annual surface reflectance composites and forest change information, our next aim was to map the land cover over the near three decade time period represented by our data cube. Knowing the issues with comparing land cover maps to relate change (see Fuller et al. 2003), we developed an approach to utilize and incorporate the disturbance data we previously generated. (BTW, Fuller et al. (2003) is a must read for those interested in land cover change; strong case made for not comparing maps to detect change, elegant statistical support).</p><p>When we use satellite data to show change between 2 time periods informative graphics can be developed that show to-from for each category, what categories are present when a disturbance occurs, or to relate the change implications to a secondary attribute (such as biomass). What happens when you have more than binary change? How can the change and transitions be shown when you have multiple years from time series analysis, say 28 years? In this post, we are aiming to get thoughts / feedback on possible opportunities for portrayal and communication of multi-year change from remote sensed data.</p><p>Here are some examples:</p><p>Binary change, to / from by category. Lots of class level transition detail. Great deal of land cover and transition information captured; Colditz et al. 2014.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CLX3IBmidJajKqxEzRf_HQ.png" /><figcaption>Gain, loss, and balance for each class, Colditz et al. 2014.</figcaption></figure><p>Here is nice example of 3 dates, 2 sets of change (Drigo et al. 2009). Change direction, area, and magnitude (via biomass). The transition to / from by area and direction is nicely captured.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/744/1*huvy3zwUebs2gEGyMaj7LA.png" /><figcaption>Pan-tropical flux diagrams of woody biomass change, by class; Drigo et al. 2009.</figcaption></figure><p>Anyone have some examples from the scientific literature or links to any scientific communication (#scicomm) pages where we might get some ideas? Please comment below, or get in touch with me, if you have any ideas. Any insights appreciated!</p><p>Reference List</p><p>Colditz, R. R., Pouliot, D., Llamas, R. M., Homer, C., Latifovic, R., Ressl, R. A., … &amp; Richardson, K. (2014). Detection of North American land cover change between 2005 and 2010 with 250m MODIS data. <em>PHOTOGRAMMETRIC ENGINEERING AND REMOTE SENSING</em>, <em>80</em>(10), 918–924.</p><p>Drigo, R., Lasserre, B., &amp; Marchetti, M. (2009). Patterns and trends in tropical forest cover. <em>Plant Biosystems</em>, <em>143</em>(2), 311–327.</p><p>Fuller, R. M., Smith, G. M., &amp; Devereux, B. J. (2003). The characterisation and measurement of land cover change through remote sensing: problems in operational applications?. <em>International Journal of Applied Earth Observation and Geoinformation</em>, <em>4</em>(3), 243–253.</p><p>Hermosilla, T., Wulder, M. A., White, J. C., Coops, N. C., &amp; Hobart, G. W. (2015a). An integrated Landsat time series protocol for change detection and generation of annual gap-free surface reflectance composites. <em>Remote Sensing of Environment</em>, <em>158</em>, 220–234. <a href="https://www.researchgate.net/publication/269113059_An_integrated_Landsat_time_series_protocol_for_change_detection_and_generation_of_annual_gap-free_surface_reflectance_composites">Link</a></p><p>Hermosilla, T., Wulder, M. A., White, J. C., Coops, N. C., &amp; Hobart, G. W. (2015). Regional detection, characterization, and attribution of annual forest change from 1984 to 2012 using Landsat-derived time-series metrics. <em>Remote Sensing of Environment</em>, <em>170</em>, 121–132. <a href="https://www.researchgate.net/publication/282327786_Regional_detection_characterization_and_attribution_of_annual_forest_change_from_1984_to_2012_using_Landsat-derived_time-series_metrics">Link</a></p><p>Hermosilla, T., Wulder, M. A., White, J. C., Coops, N. C., Hobart, G. W., &amp; Campbell, L. B. (2016). Mass data processing of time series Landsat imagery: pixels to data products for forest monitoring. <em>International Journal of Digital Earth</em>, <em>9</em>(11), 1035–1054. <a href="http://www.tandfonline.com/doi/full/10.1080/17538947.2016.1187673">Link</a></p><p>White, J. C., Wulder, M. A., Hobart, G. W., Luther, J. E., Hermosilla, T., Griffiths, P., … &amp; Guindon, L. (2014). Pixel-based image compositing for large-area dense time series applications and science. <em>Canadian Journal of Remote Sensing</em>, <em>40</em>(3), 192–212. <a href="https://www.researchgate.net/publication/267509030_Pixel-Based_Image_Compositing_for_Large-Area_Dense_Time_Series_Applications_and_Science">Link</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=32a20e95f76" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Using lidar to select yield curves through template matching for forest growth and volume…]]></title>
            <link>https://medium.com/@mikewulder/using-lidar-to-select-yield-curves-through-template-matching-for-forest-growth-and-volume-1f5253150c98?source=rss-39acee693c3a------2</link>
            <guid isPermaLink="false">https://medium.com/p/1f5253150c98</guid>
            <category><![CDATA[forest]]></category>
            <category><![CDATA[forestry]]></category>
            <category><![CDATA[environment]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[lidar]]></category>
            <dc:creator><![CDATA[Mike Wulder]]></dc:creator>
            <pubDate>Wed, 22 Feb 2017 23:47:54 GMT</pubDate>
            <atom:updated>2017-02-22T23:47:54.017Z</atom:updated>
            <content:encoded><![CDATA[<h3>Using lidar to select yield curves through template matching for forest growth and volume estimation</h3><p>In a recent paper led by <a href="https://twitter.com/PiotrTompalski">Dr. Piotr Tompalski</a>, of the <a href="https://twitter.com/IRSS_UBC">Integrated Remote Sensing Studio (IRSS)</a> at the University of British Columbia, we communicated a concept of template matching for estimating forest growth and volume. In this post I ask Dr. Tompalski a few questions with an aim to make clear the concept of template matching and how we used it to produce lidar-informed estimates of forest growth and volume.</p><blockquote>What are the benefits of a template matching approach? Tompalski replies that “…the benefits of this approach are threefold: increased spatial level of detail of the growth and yield analysis, optimization of the yield curve selection, and the information on the uncertainty of curve selection”.</blockquote><p>By way of background, information on forest growth is crucial for sustainable forest management. This information determines many important management decisions like optimal harvest age or silvicultural practices. To predict future stand attributes foresters use growth models. These models, based on the input information on current forest stand conditions like age, species composition, and site index, “grow” stand into future by specified age increments. Stand level forest inventory is a sufficient source of input information, although growth models can also alter the growth with additional information on disturbances, fertilization, or thinning. Foresters can therefore perform various simulations of future stand conditions and observe outcomes of their decisions.</p><p>Typically, a yield curve for a single stand generated with a growth model contains information on forest merchantable and total volume, top and mean height, basal area, mean diameter, and number of trees, for each age increment. An example of growth curves for four different stands is shown below. Four different curves represent change in time (0 to 200 years) of three stand attributes — height (H, meters), number of trees (N, trees/ha), and total volume (V, m3/ha). These curves demonstrate how stands composed of different species and representing different levels of productivity change through time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*ONQkKKbO6WsFel46FVrdCg.png" /><figcaption>Relationship between Height (H), Number of trees per hectare (N), and total volume (V).</figcaption></figure><p>Foresters use polygon-level inventory data to manage forest stands. These data include information on species composition, age, height, site index, basal area, total volume, and other stand attributes, at current moment in time. Because of large areas of these polygons (often &gt;100 ha), the inventory information represents the averages (e.g. height) or sums (e.g. basal area) of stand attributes. The within-stand variability is therefore not preserved, i.e. volume or basal area of a particular subset of the stand is not known. Consequently, the growth and yield analysis based on such input data will also lack the detail on the stand attributes within the stand.</p><p>However, the polygon-level forest inventory can be enhanced and the level of detail improved. Airborne laser scanning (ALS) can be used to estimate forest stand attributes. As proven by a large number of scientific publications, ALS-based metrics that characterize 3D structure of forest stands, can be used to predict height, diameter, basal area, volume, or biomass of a stand. The method of estimating forest stand attributes, like basal area or volume, can be based on pixels, or cells, that are typically 20–30 m on a side. The predictive models are built that estimate forest attribute (e.g. volume) based on ALS-based metrics calculated for each cell. Such an approach, referred to as an “area-based approach” (ABA) is popular and used operationally over a range of forest environments.</p><p>While ALS proved to be an excellent tool to characterize tree height and a number of important forest stand attributes, it is currently not successful at extracting species or stand age. These two pieces of information are the crucial inputs for forest growth models. Therefore the cell-level stand attributes derived with ALS have to be supported by information from polygon-level forest inventory.</p><p>However, as Tompalski notes, “The process of assigning a yield curve to each cell is more complicated compared to a traditional situation when stand-level input information is used. To introduce more spatial detail of the growth and yield analysis, we perform additional steps that allow us to select the best matching curve to each cell, assess the level of uncertainty”. Below, the process is described in detail.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/266/1*NHvHihMpoR6EtRA7E-V_lg.png" /></figure><p>Tompalski explains that:</p><p>“We start by generating a growth curve template database that represents all growth scenarios of forest stands. The database is generated with an existing growth model and is based on a range of the required input parameters (e.g. species, site index, age). Those input parameters are obtained from existing forest inventory data.”</p><p>Next, we use ALS-derived stand attributes and stand age from the inventory data. The ALS-derived attributes are raster layers with pixel size of 20 m, and therefore contain much more detail on forest attributes. At this stage information on stand age is also imported and assigned into each of the 20 m cells.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/477/1*Pcd_wwQmYawKVIMkW-AySA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/245/1*ksIyLLv9pl-rWTLNzWrQIA.png" /></figure><p>To find a matching yield curve in the template database we look for a curve that has a given value of one of the attributes at given age. For each of the ALS-derived attributes (HMAX, HL, QMD, and V in this case), we identify a yield curve with the closest value.</p><p>Because there are discrepancies between the existing stand conditions and the attributes derived with the growth model, and the models used to create the ALS cell-level estimates are not error-free, the curve matching based on each of the ALS-predicted attributes may result with a different curve. In our case, with four ALS-derived attributes, four candidate curves are identified. In some cases the same curve may be matched based on different input attributes. However, very often the matched candidate curves differ as shown on the image below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/388/1*_T4IpqTibcVELCDc3UbpAA.png" /></figure><p>So, now you have all these yield curves, how do you choose the right one?</p><p>Tompalski replies that, “The final curve is assigned by averaging the four candidate curve. In this step we take into account the known accuracy of the models used to estimate the ALS-derived attributes. That is, we know that the accuracy of the model predicting stand top height will be more accurate than the model predicting stand volume. We can therefore use a weighted mean and assign different weights to each of the candidate curves, based on the coefficient of determination of each of the four predictive models. In addition, we can calculate the level of uncertainty that informs on the discrepancies between the candidate curves.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/426/1*tNwTf3G7YKyz2ueGZCQlpQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/189/1*90N-3SbYQR_9CJcrl6Yn1Q.png" /></figure><p>At this stage a yield curve is assigned to the input cell. It can be used to project cell attributes to a desired age. In our case we projected the attributes to the typical rotation age in the study area (80 years), and generated raster layers representing growth model outputs for each of the cells. Note that the number of output layers exceeds the number of rasters used as input. This is because the growth models very often generate stand characteristics based on other parameters. In this case both the number of trees (TPH) and basal area (BA) were not used as inputs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*lHXfPbEx43JM6pAum9NllA.png" /></figure><p>What are the benefits of such an approach? Tompalski indicates that “…the benefits of this approach are threefold: increased spatial level of detail of the growth and yield analysis, optimization of the yield curve selection, and the information on the uncertainty of curve selection. The template matching approach allows us to refine the selection of appropriate growth and yield curve, while also informing on the error and alternate growth trajectories based upon input forest characteristics.”</p><p><strong>Reference</strong>:</p><p>Tompalski, P., N.C. Coops, J.C. White, and M.A. Wulder, 2016. Enhancing Forest Growth and Yield Predictions with Airborne Laser Scanning Data: Increasing Spatial Detail and Optimizing Yield Curve Selection through Template Matching. Forests. 7(12):255.</p><p>Open Access link: DOI:<a href="http://www.mdpi.com/1999-4907/7/11/255">10.3390/f7110255</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1f5253150c98" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>