#Lidar is an operational technology for #forestinventory. Single-photon lidar (#SPL) may provide an acquisition advantage for large areas. How does #SPL perform in an area-based #forest inventory? Some things we have learned (i.e. a thread)!
2/ We used #SPL and 269 field plots to generate an area-based operational #forestinventory over a 15000 ha study area of temperate mixedwood forest representing a complex assemblage of tree species, forest structures, and management histories. bit.ly/37GTeZG
3/ We validated our #forestinventory estimates at the stand-level using independent field measured data. Why? Because the stand is the fundamental spatial unit for forest management and planning, and the unit at which decisions are made. #CFSEFIbit.ly/37GTeZG
4/ We predicted a suite of #forestinventory attributes and assessed model performance overall and by forest type. bit.ly/37GTeZG
5/ Accuracy and bias varied by forest type. Stand-level validation was important for assessing model performance in different stand conditions. Overall, plot-level results (OOB) overestimated model error and underestimated bias relative to the stand-level validation results.
6/ #Forestinventory attributes such as volume and basal area were most challenging in managed pine stands and red pine plantations, which have very different stand structures. bit.ly/37GTeZG
7/ Results were on par with area-based outcomes achieved using linear-mode #lidar in this same study area, as well as with results reported in other studies using SPL data (in boreal forests): bit.ly/3u0dhLE and bit.ly/2PyHpPy.
8/ For large FMAs, the ability to acquire #lidar data from a single instrument with a consistent specification can be critical for operational implementation. From a model performance perspective, SPL data provided accurate #forestinventory attribute estimates with minimal bias.
10/ To learn more about our project "Exploring the innovation potential of single photon lidar for Ontario's eFRI" (including recorded webinar): bit.ly/32ReN72 and also bit.ly/3e134cE
11/ It is worth noting that although #SPL offers an advantage for #lidar data acquisition over large areas because of the high altitude used for acquisition (~3800 m agl and as documented bit.ly/3u0dhLE and bit.ly/2PyHpPy), low clouds/smoke can be problematic.
12/ Finally, to access the SPL and field plot data used in our research, as well as the predicted EFI attribute layers, visit the PRF remote sensing supersite [bit.ly/3vqaADD] and use search terms PRF and SPL. #PRFSS#CFSEFI #OpenAccess: bit.ly/37GTeZG
• • •
Missing some Tweet in this thread? You can try to
force a refresh
An increasingly common question: “We generated an enhanced forest inventory with #lidar data and ground plots for one of our management units. Can we use the same models for our other management areas?” #CFSEFI#forestinventory
A great question and one that is top of mind for many of us who generate and use #forestinventory data. So what does the #science tell us about the transferability of area-based models? #CFSEFI
Fekety et al. 2015: "Mapping response variables at the landscape level demonstrates that the relationship between field data and LiDAR metrics holds true even though the data were collected in different years."