Testing a Landsat-based approach for mapping disturbance causality in U.S. forests

Schroeder, Todd A.; Schleeweis, Karen; Moisen, Gretchen G.; Toney, Chris; Cohen, Warren B.; Freeman, Elizabeth A.; Yang, Zhiqiang; Huang, Chengquan

doi:10.1016/j.rse.2017.03.033

Cited by 61 publications

(53 citation statements)

References 66 publications

(72 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Like Schroeder et al (2017) both our disturbed classes have recovery trajectories that resemble a 'vee' shape [40]. However, even though the shape is similar, we showed that development disturbance can be separated from non-development disturbances using measurable differences in the Recovery Slope (low to high) and Recovery Maximum (High), focusing solely on post-disturbance (aka "post-rate") spectral change.…”

Section: Discussionsupporting

confidence: 69%

“…However, even though the shape is similar, we showed that development disturbance can be separated from non-development disturbances using measurable differences in the Recovery Slope (low to high) and Recovery Maximum (High), focusing solely on post-disturbance (aka "post-rate") spectral change. Schroeder et al also found that the ability to detect forest conversion was quite variable (error rates: 0-77%) across test scenes, with a relative dearth of reference data a major issue [40]. Nonetheless, like our study, they were able to separate conversion from other anthropogenic disturbances such as forest harvesting, but the specific utility of their algorithm for detecting low density forest disturbances is unknown.…”

Section: Discussionmentioning

confidence: 45%

“…Nationally, disturbance events impact <3% of forest area per year, making any specific disturbance type rare [40,44]. As we noted earlier, Schroeder et al also allude to the difficulty in collecting enough reference data to sufficiently capture disturbances like low density development [40].…”

Section: Discussionmentioning

confidence: 93%

“…As we noted earlier, Schroeder et al also allude to the difficulty in collecting enough reference data to sufficiently capture disturbances like low density development [40]. Through the use of well permit data and housing start data we had 750 k points from which we could sample, in addition to the random points generated for the other disturbance types.…”

Section: Discussionmentioning

confidence: 99%

See 3 more Smart Citations

Identifying Forest Impacted by Development in the Commonwealth of Virginia through the Use of Landsat and Known Change Indicators

House

Wynne

2018

Remote Sensing

View full text Add to dashboard Cite

This study examines the effectiveness of using the Normalized Difference Vegetation Index (NDVI) derived from 1326 different Landsat Thematic Mapper and Enhanced Thematic Mapper images in finding low density development within the Commonwealth of Virginia's forests. Individual NDVI images were stacked by year for the years 1995-2011 and the yearly maximum for each pixel was extracted, resulting in a 17-year image stack of all yearly maxima (a 98.7% data reduction). Using location data from housing starts and well permits, known previously forested housing starts were isolated from all other forest disturbance types. Samples from development disturbances and other forest disturbances, as well as from undisturbed forest, were used to derive vegetation index thresholds enabling separation of disturbed forest from undisturbed forest. Disturbances, once identified, could be separated into Development Disturbances and Non-Development Disturbances using a classification tree and only two variables from the Disturbance Detection and Diagnostics (D 3 ) algorithm: the maximum NDVI in the available recovery period and the slope between the NDVI value at the time of the disturbance and the maximum NDVI in the available recovery period. Low density development disturbances of previous forest land cover had an F-measure, combining precision and recall into a single class-specific accuracy (β = 1), of 0.663. We compared our results to the NLCD 2001-2011 land cover changes from any forest (classes 41, 42, 43, and 90) to any developed (classes 21, 22, 23, and 24), resulting in an F-measure of 0.00 for the same validation points. Landsat time series stacks thus show promise for identifying even the small changes associated with low density development that have been historically overlooked/underestimated by prior mapping efforts. However, further research is needed to ensure that (1) the approach will work in other forest biomes and (2) enabling detection of these important, but spatially and spectrally subtle, disturbances still ensures accurate detection of other forest disturbances.

show abstract

Section: Discussionsupporting

confidence: 69%

Section: Discussionmentioning

confidence: 45%

Section: Discussionmentioning

confidence: 93%

Section: Discussionmentioning

confidence: 99%

See 2 more Smart Citations

Identifying Forest Impacted by Development in the Commonwealth of Virginia through the Use of Landsat and Known Change Indicators

House

Wynne

2018

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Remote sensing techniques have been widely used in forestry to map and monitor forest dynamics (Cohen et al, 2016;Hansen et al, 2013;Schroeder et al, 2011;White et al, 2016White et al, , 2017. The applications of optical satellite imagery in forestry have increased dramatically at both temporal and spatial scales, particularly in the recent decade, with open access to Landsat satellite imagery (Schroeder et al, 2017;Vogelmann et al, 2017;Wang et al, 2016;White et al, 2017;Wulder et al, 2012a). A number of satellite imagery-derived vegetation indices, such as normalized difference vegetation index (NDVI), normalized difference water index (NDWI), and normalized burn ratio (NBR) have been developed to monitor forest dynamics.…”

Section: Introductionmentioning

confidence: 99%

Evaluating the uncertainty of Landsat-derived vegetation indices in quantifying forest fuel treatments using bi-temporal LiDAR data

Lai-ping

et al. 2018

Ecological Indicators

View full text Add to dashboard Cite

A B S T R A C TForest ecosystems in the American west have long been influenced by timber harvests and fire suppression, and recently through treatments that reduce fuel for fire management. Precisely quantifying the structural changes to forests caused by fuel treatments is an essential step to evaluate their impacts. Satellite imagery-derived vegetation indices, such as the normalized difference vegetation index (NDVI), have been widely used to map forest dynamics. However, uncertainties in using these vegetation indices to quantify forest structural changes have not been thoroughly studied, mainly due to the lack of wall-to-wall validation data. In this study we generated forest structural changes in aboveground biomass (AGB) and canopy cover as a result of fuel treatments using bitemporal airborne light detection and ranging (LiDAR) data and field measurements in a mixed coniferous forest of northern Sierra Nevada, California, USA. These LiDAR-derived forest structural measures were used to evaluate the uncertainties of using Landsat-derived vegetation indices to quantify treatments. Our results confirmed that vegetation indices can accurately map the extents of forest disturbance and canopy cover changes caused by fuel treatments, but the accuracy in quantifying AGB changes varied by the pre-treatment forest densities and treatment intensity. Changes in vegetation indices had relatively weaker correlations (coefficient of determination < 0.45) to biomass changes in forests with sparse (AGB < 100 Mg/ha) or dense biomass (AGB > 700 Mg/ha), than in forests with moderate-density (AGB between 100 Mg/ha and 700 Mg/ha) before the disturbances. Moreover, understory treatments (canopy height < 10 m) were poorly indicated by changes in satellite-derived vegetation indices. Our results suggest that when relating vegetation indices to AGB changes, researchers and managers should be cautious about their uncertainties in extremely dense or sparse forests, particularly when treatments mainly removed small trees or understory fuels. Tempel et al., 2014). Because these management practices are so impactful on forest structure, accurate measurement of the changes to forest structure as a result of fuel treatments is necessary, but doing so remains challenging. Accurate and timely quantification of forest changes in abundance, production, and spatial pattern is a necessary step for forest fuel treatment evaluation (Huang et al., 2009;Su et al., 2016a). Traditional (M. Kelly).Ecological Indicators 95 (2018) 298-310 1470-160X/

show abstract

Optical Satellite Image Time Series Analysis for Environment Applications: From Classical Methods to Deep Learning and Beyond

Molinier¹,

Miettinen²,

Ienco³

et al. 2021

Change Detection and Image Time Series Analysis 2

View full text Add to dashboard Cite

Testing a Landsat-based approach for mapping disturbance causality in U.S. forests

Cited by 61 publications

References 66 publications

Identifying Forest Impacted by Development in the Commonwealth of Virginia through the Use of Landsat and Known Change Indicators

Identifying Forest Impacted by Development in the Commonwealth of Virginia through the Use of Landsat and Known Change Indicators

Evaluating the uncertainty of Landsat-derived vegetation indices in quantifying forest fuel treatments using bi-temporal LiDAR data

Optical Satellite Image Time Series Analysis for Environment Applications: From Classical Methods to Deep Learning and Beyond

Contact Info

Product

Resources

About