@article{Kukunda2019Scale,
 abstract = {Heterogeneity in forest structure, naturally occurring or induced by management, is continuous in space and time. However, measures used to quantify structure of forests are scale-variant, as they rely on bounded observations on either ecological or forest inventory observation units. The understanding of the influence of the scale of observation in mapping of forest structural heterogeneity is limited. Therefore, we researched into effects of plot size on quantifying forest structural heterogeneity, where we describe heterogeneity by three indices in stands under different management systems. In addition, we studied the performance of structural indices in separating different forest management systems across plot sizes, and created wall to wall maps of the indices using airborne LiDAR metrics describing the vertical distribution of canopy heights at different scales of observation. The studied indices are: Gini Coefficient (GC), Structural Complexity Index (SCI), and Enhanced Structural Complexity Index (ESCI). SCI and ESCI require fully mapped plots whereas GC has no information on individual tree locations. Inventory data from 95 one-hectare plots covering a range of management intensities from un-managed to age class forests were used. We quantified the three structural indices for 18 plot sizes ranging from 225 m2 to 10,000 m2. Linear fixed effects models were used to study the effects of plot sizes in different levels of structural heterogeneity and Random Forest (RF) models used to provide wall-to-wall maps at varying scales from airborne LiDAR data. The simulation showed that all indices were influenced by the scale of observation with larger effects for plots in forests with higher structural heterogeneity. For the data analyzed we found a threshold scale for enumerating stand structural heterogeneity between 900 m2 and 2500 m2. However, stable field and remote sensing predictions of stand structural heterogeneity required plots at least ⩾2500 m2. Compared to GC, SCI and ESCI improved separation of forest structure in the three management systems and at all observed scales. A change of plot sizes affected bivariate relationships between structural indices and airborne LiDAR metrics as well as the resultant predictive models. Smaller plot sizes yielded weaker relationships and predictive models. All structure indices were predicted from airborne LiDAR with RMSE⩽22{\%} at scales equal or larger than the identified threshold plot size. These findings are relevant to optimize plot sizes for efficient inventory and mapping of forest structural heterogeneity, as well as for the design of natural resource inventories. Additionally, derived maps are useful for studies on forest structure and the link with forest growth, degradation, management intensity, productivity, and biodiversity in the regions.},
 author = {Kukunda, Collins B. and Becksch{\"a}fer, Philip and Magdon, Paul and Schall, Peter and Wirth, Christian and Kleinn, Christoph},
 year = {2019},
 title = {Scale-guided mapping of forest stand structural heterogeneity from airborne {LiDAR}},
 keywords = {ecol;phd},
 pages = {410--425},
 volume = {102},
 issn = {1470160X},
 journal = {Ecological Indicators},
 doi = {10.1016/j.ecolind.2019.02.056},
 howpublished = {refereed}
}

{"_id":"K8dkjf5F6cgtd2KuE","bibbaseid":"kukunda-beckschfer-magdon-schall-wirth-kleinn-scaleguidedmappingofforeststandstructuralheterogeneityfromairbornelidar-2019","downloads":0,"creationDate":"2019-03-06T10:04:48.389Z","title":"Scale-guided mapping of forest stand structural heterogeneity from airborne LiDAR","author_short":["Kukunda, C. B.","Beckschäfer, P.","Magdon, P.","Schall, P.","Wirth, C.","Kleinn, C."],"year":2019,"bibtype":"article","biburl":"http://www.uni-goettingen.de/de/document/download/9d7c40531010bf5be953ccd9446e47ae.bib/GRK1644BibHomepage.bib","bibdata":{"bibtype":"article","type":"article","abstract":"Heterogeneity in forest structure, naturally occurring or induced by management, is continuous in space and time. However, measures used to quantify structure of forests are scale-variant, as they rely on bounded observations on either ecological or forest inventory observation units. The understanding of the influence of the scale of observation in mapping of forest structural heterogeneity is limited. Therefore, we researched into effects of plot size on quantifying forest structural heterogeneity, where we describe heterogeneity by three indices in stands under different management systems. In addition, we studied the performance of structural indices in separating different forest management systems across plot sizes, and created wall to wall maps of the indices using airborne LiDAR metrics describing the vertical distribution of canopy heights at different scales of observation. The studied indices are: Gini Coefficient (GC), Structural Complexity Index (SCI), and Enhanced Structural Complexity Index (ESCI). SCI and ESCI require fully mapped plots whereas GC has no information on individual tree locations. Inventory data from 95 one-hectare plots covering a range of management intensities from un-managed to age class forests were used. We quantified the three structural indices for 18 plot sizes ranging from 225 m2 to 10,000 m2. Linear fixed effects models were used to study the effects of plot sizes in different levels of structural heterogeneity and Random Forest (RF) models used to provide wall-to-wall maps at varying scales from airborne LiDAR data. The simulation showed that all indices were influenced by the scale of observation with larger effects for plots in forests with higher structural heterogeneity. For the data analyzed we found a threshold scale for enumerating stand structural heterogeneity between 900 m2 and 2500 m2. However, stable field and remote sensing predictions of stand structural heterogeneity required plots at least ⩾2500 m2. Compared to GC, SCI and ESCI improved separation of forest structure in the three management systems and at all observed scales. A change of plot sizes affected bivariate relationships between structural indices and airborne LiDAR metrics as well as the resultant predictive models. Smaller plot sizes yielded weaker relationships and predictive models. All structure indices were predicted from airborne LiDAR with RMSE⩽22% at scales equal or larger than the identified threshold plot size. These findings are relevant to optimize plot sizes for efficient inventory and mapping of forest structural heterogeneity, as well as for the design of natural resource inventories. Additionally, derived maps are useful for studies on forest structure and the link with forest growth, degradation, management intensity, productivity, and biodiversity in the regions.","author":[{"propositions":[],"lastnames":["Kukunda"],"firstnames":["Collins","B."],"suffixes":[]},{"propositions":[],"lastnames":["Beckschäfer"],"firstnames":["Philip"],"suffixes":[]},{"propositions":[],"lastnames":["Magdon"],"firstnames":["Paul"],"suffixes":[]},{"propositions":[],"lastnames":["Schall"],"firstnames":["Peter"],"suffixes":[]},{"propositions":[],"lastnames":["Wirth"],"firstnames":["Christian"],"suffixes":[]},{"propositions":[],"lastnames":["Kleinn"],"firstnames":["Christoph"],"suffixes":[]}],"year":"2019","title":"Scale-guided mapping of forest stand structural heterogeneity from airborne LiDAR","keywords":"ecol;phd","pages":"410–425","volume":"102","issn":"1470160X","journal":"Ecological Indicators","doi":"10.1016/j.ecolind.2019.02.056","howpublished":"refereed","bibtex":"@article{Kukunda2019Scale,\r\n abstract = {Heterogeneity in forest structure, naturally occurring or induced by management, is continuous in space and time. However, measures used to quantify structure of forests are scale-variant, as they rely on bounded observations on either ecological or forest inventory observation units. The understanding of the influence of the scale of observation in mapping of forest structural heterogeneity is limited. Therefore, we researched into effects of plot size on quantifying forest structural heterogeneity, where we describe heterogeneity by three indices in stands under different management systems. In addition, we studied the performance of structural indices in separating different forest management systems across plot sizes, and created wall to wall maps of the indices using airborne LiDAR metrics describing the vertical distribution of canopy heights at different scales of observation. The studied indices are: Gini Coefficient (GC), Structural Complexity Index (SCI), and Enhanced Structural Complexity Index (ESCI). SCI and ESCI require fully mapped plots whereas GC has no information on individual tree locations. Inventory data from 95 one-hectare plots covering a range of management intensities from un-managed to age class forests were used. We quantified the three structural indices for 18 plot sizes ranging from 225 m2 to 10,000 m2. Linear fixed effects models were used to study the effects of plot sizes in different levels of structural heterogeneity and Random Forest (RF) models used to provide wall-to-wall maps at varying scales from airborne LiDAR data. The simulation showed that all indices were influenced by the scale of observation with larger effects for plots in forests with higher structural heterogeneity. For the data analyzed we found a threshold scale for enumerating stand structural heterogeneity between 900 m2 and 2500 m2. However, stable field and remote sensing predictions of stand structural heterogeneity required plots at least ⩾2500 m2. Compared to GC, SCI and ESCI improved separation of forest structure in the three management systems and at all observed scales. A change of plot sizes affected bivariate relationships between structural indices and airborne LiDAR metrics as well as the resultant predictive models. Smaller plot sizes yielded weaker relationships and predictive models. All structure indices were predicted from airborne LiDAR with RMSE⩽22{\\%} at scales equal or larger than the identified threshold plot size. These findings are relevant to optimize plot sizes for efficient inventory and mapping of forest structural heterogeneity, as well as for the design of natural resource inventories. Additionally, derived maps are useful for studies on forest structure and the link with forest growth, degradation, management intensity, productivity, and biodiversity in the regions.},\r\n author = {Kukunda, Collins B. and Becksch{\\\"a}fer, Philip and Magdon, Paul and Schall, Peter and Wirth, Christian and Kleinn, Christoph},\r\n year = {2019},\r\n title = {Scale-guided mapping of forest stand structural heterogeneity from airborne {LiDAR}},\r\n keywords = {ecol;phd},\r\n pages = {410--425},\r\n volume = {102},\r\n issn = {1470160X},\r\n journal = {Ecological Indicators},\r\n doi = {10.1016/j.ecolind.2019.02.056},\r\n howpublished = {refereed}\r\n}\r\n\r\n\r\n","author_short":["Kukunda, C. B.","Beckschäfer, P.","Magdon, P.","Schall, P.","Wirth, C.","Kleinn, C."],"key":"Kukunda2019Scale","id":"Kukunda2019Scale","bibbaseid":"kukunda-beckschfer-magdon-schall-wirth-kleinn-scaleguidedmappingofforeststandstructuralheterogeneityfromairbornelidar-2019","role":"author","urls":{},"keyword":["ecol;phd"],"metadata":{"authorlinks":{}},"downloads":0},"search_terms":["scale","guided","mapping","forest","stand","structural","heterogeneity","airborne","lidar","kukunda","beckschäfer","magdon","schall","wirth","kleinn"],"keywords":["ecol;phd"],"authorIDs":[],"dataSources":["psxr4mFyE5JDwFLuZ","2w3D54bmLuhpt4TNv","t8S6Y6RWEwDAQHiSQ","cLGdYAfLyvQDgrYmh"]}