Characterizing stand-level forest canopy cover and height using Landsat time series, samples of airborne LiDAR, and the Random Forest algorithm

Ahmed, Oumer S.; Franklin, Steven E.; Wulder, Michael A.; White, Joanne C.

doi:10.1016/j.isprsjprs.2014.11.007

Cited by 149 publications

(89 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Canopy height derived from airborne Lidar was correlated with both TCT variables and disturbance information obtained from a trajectory-based disturbance characterization method (Table 3), and a strong positive correlation is found between Lidar-derived MCH and the TCA. Our study has confirmed that TCA is particularly a well suited index to long-term change detection studies that include MSS data (e.g., [18,30]). Several other spectral variables or indices can also be used for disturbance characterization especially when MSS data are not included in the time series.…”

Section: Discussionsupporting

confidence: 77%

“…The random selection of independent variables is performed at each level of the tree. A comprehensive explanation of the algorithm can be found in Breiman [63,64] and its application for forest parameter estimation can be found in Hudak et al [42], McInerney and Nieuwenhuis [65], Powell et al [30], Gleason and Im [66] and Ahmed et al [18]. The RF modeling was performed using the "random forest" package [67] in R statistical language [61].…”

Section: Mean Canopy Height Estimation Using Multiple Regression and mentioning

confidence: 99%

“…However, wall-to-wall acquisitions of Lidar data remain cost prohibitive for large forest areas or in areas without a competitive commercial Lidar data acquisition community. In an effort to overcome cost limitation of Lidar data, recent studies report on the integration of Lidar data with optical satellite remotely sensed data to estimate canopy structure from a relatively small area covered by Lidar sample over a larger area covered by the optical scene [18][19][20][21]. These studies mainly focused on managed temperate forests and to date there is a paucity of such studies in unmanaged tropical forests.…”

Section: Introductionmentioning

confidence: 99%

“…For example, Pflugmacher et al [35] demonstrated the inclusion of disturbance and recovery metrics derived from spectral profiles of annual Landsat time series provided an improved prediction of current forest structural attributes. Ahmed et al [18] suggested that pre-stratification of the forest types based on the disturbance date and the addition of time since disturbance information improved forest canopy cover and height estimation in a managed temperate forest. Li et al [36] showed that inclusion of disturbance information characterized by vegetation change tracker algorithm yielded the lowest validation error for the estimation of young forest height calculated from space borne LiDAR.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Estimation of Airborne Lidar-Derived Tropical Forest Canopy Height Using Landsat Time Series in Cambodia

et al. 2014

View full text Add to dashboard Cite

Abstract:In this study, we test and demonstrate the utility of disturbance and recovery information derived from annual Landsat time series to predict current forest vertical structure (as compared to the more common approaches, that consider a sample of airborne Lidar and single-date Landsat derived variables). Mean Canopy Height (MCH) was estimated separately using single date, time series, and the combination of single date and time series variables in multiple regression and random forest (RF) models. The combination of single date and time series variables, which integrate disturbance history over the entire time series, overall provided better MCH prediction than using either of the two sets of variables separately. In general, the RF models resulted in improved performance in all estimates over those using multiple regression. The lowest validation error was obtained using Landsat time series variables in a RF model (R 2 = 0.75 and RMSE = 2.81 m). Combining single date and time series data was more effective when the RF model was used (opposed to multiple regression). The RMSE for RF mean canopy height prediction was reduced by 13.5% when combining the two sets of variables as compared to the 3.6% RMSE decline presented by multiple regression. This study demonstrates the value of airborne Lidar and long term Landsat observations to generate estimates of forest canopy height using the random forest algorithm.

show abstract

Section: Discussionsupporting

confidence: 77%

Section: Mean Canopy Height Estimation Using Multiple Regression and mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Estimation of Airborne Lidar-Derived Tropical Forest Canopy Height Using Landsat Time Series in Cambodia

et al. 2014

View full text Add to dashboard Cite

show abstract

“…Nonetheless, our results are similar to previous stand-level height estimation. For example, Ahmed et al [49] modeled forest heights at the stand level with Landsat time series images and airborne LiDAR data using multiple regression and RF. Their RF models resulted in R 2 of 0.88, RMSE of 2.39 m for mature forests, R 2 of 0.79, RMSE of 3.52 m for young forests, and R 2 of 0.82, RMSE of 3.17 m for combined forests.…”

Section: Validation Of Forest Stand Height Modelsmentioning

confidence: 99%

Machine Learning Approaches for Estimating Forest Stand Height Using Plot-Based Observations and Airborne LiDAR Data

Lee

Kim

et al. 2018

Forests

View full text Add to dashboard Cite

Effective sustainable forest management for broad areas needs consistent country-wide forest inventory data. A stand-level inventory is appropriate as a minimum unit for local and regional forest management. South Korea currently produces a forest type map that contains only four categorical parameters. Stand height is a crucial forest attribute for understanding forest ecosystems that is currently missing and should be included in future forest type maps. Estimation of forest stand height is challenging in South Korea because stands exist in small and irregular patches on highly rugged terrain. In this study, we proposed stand height estimation models suitable for rugged terrain with highly mixed tree species. An arithmetic mean height was used as a target variable. Plot-level height estimation models were first developed using 20 descriptive statistics from airborne Light Detection and Ranging (LiDAR) data and three machine learning approaches-support vector regression (SVR), modified regression trees (RT) and random forest (RF). Two schemes (i.e., central plot-based (Scheme 1) and stand-based (Scheme 2)) for expanding from the plot level to the stand level were then investigated. The results showed varied performance metrics (i.e., coefficient of determination, root mean square error, and mean bias) by model for forest height estimation at the plot level. There was no statistically significant difference among the three mean plot height models (i.e., SVR, RT and RF) in terms of estimated heights and bias (p-values > 0.05). The stand-level validation based on all tree measurements for three selected stands produced varied results by scheme and machine learning used. It implies that additional reference data should be used for a more thorough stand-level validation to identify statistically robust approaches in the future. Nonetheless, the research findings from this study can be used as a guide for estimating stand heights for forests in rugged terrain and with complex composition of tree species.

show abstract

Monitoring tropical forest succession at landscape scales despite uncertainty in Landsat time series

Caughlin

Barber

Asner

et al. 2020

Ecological Applications

View full text Add to dashboard Cite

Forecasting rates of forest succession at landscape scales will aid global efforts to restore tree cover to millions of hectares of degraded land. While optical satellite remote sensing can detect regional land cover change, quantifying forest structural change is challenging. We developed a state-space modeling framework that applies Landsat satellite data to estimate variability in rates of natural regeneration between sites in a tropical landscape. Our models work by disentangling measurement error in Landsat-derived spectral reflectance from process error related to successional variability. We applied our modeling framework to rank rates of forest succession between 10 naturally regenerating sites in Southwestern Panama from about 2001 to 2015 and tested how different models for measurement error impacted forecast accuracy, ecological inference, and rankings of successional rates between sites. We achieved the greatest increase in forecasting accuracy by adding intra-annual phenological variation to a model based on Landsatderived normalized difference vegetation index (NDVI). The best-performing model accounted for inter-and intra-annual noise in spectral reflectance and translated NDVI to canopy height via Landsat-lidar fusion. Modeling forest succession as a function of canopy height rather than NDVI also resulted in more realistic estimates of forest state during early succession, including greater confidence in rank order of successional rates between sites. These results establish the viability of state-space models to quantify ecological dynamics from time series of space-borne imagery. State-space models also provide a statistical approach well-suited to fusing high-resolution data, such as airborne lidar, with lower-resolution data that provides better temporal and spatial coverage, such as the Landsat satellite record. Monitoring forest succession using satellite imagery could play a key role in achieving global restoration targets, including identifying sites that will regain tree cover with minimal intervention.

show abstract

Characterizing stand-level forest canopy cover and height using Landsat time series, samples of airborne LiDAR, and the Random Forest algorithm

Cited by 149 publications

References 66 publications

Estimation of Airborne Lidar-Derived Tropical Forest Canopy Height Using Landsat Time Series in Cambodia

Estimation of Airborne Lidar-Derived Tropical Forest Canopy Height Using Landsat Time Series in Cambodia

Machine Learning Approaches for Estimating Forest Stand Height Using Plot-Based Observations and Airborne LiDAR Data

Monitoring tropical forest succession at landscape scales despite uncertainty in Landsat time series

Contact Info

Product

Resources

About