Reliable tree canopy cover (TCC) products are vital for national forest inventory, land process modeling andforest dynamic monitoring. The new generation of space-based laser altimeter, GEDI, offers a three-dimensional(3D) insight on the forest structure, shaping the paradigm of structural variable estimation. However, the generality of newly released GEDI level-2 TCC product version 2 was less investigated across various forest types.Additionally, satellite-derived product validation usually suffers from the geolocation mismatch between satelliteand reference data. In this study, we comprehensively validated the GEDI TCC product across seven forest typesusing the reference TCC derived from several public and private aerial LiDAR datasets after geographicalregistration, and crossly compared with a commonly-used passive satellite product (i.e., GFCC TCC). As thereference aerial TCC maps were derived using various aerial LiDAR instruments, we investigated the consistencyof TCC estimation among them using simulation datasets and found that the distributions of TCC relative bias(biasR, %) were almost identical and the differences of relative RMSE (rRMSE, %) was less than 0.2%. Throughthe registration process, we found that the geolocation offsets of GEDI footprints tended to be independent ofazimuth directions and their average was about 10 m, verifying the necessity of registration during the validationprocess. Importantly, the post-registration validation of GEDI TCC showed an average RMSE of 0.10 and anaverage R2 of 0.85 for all forest types, resulting in a decrease of RMSE of up to 0.15 and an increase of R2 of up to0.33 compared to the pre-registration validation. The inter-comparison also exhibited improved consistencybetween GEDI and GFCC TCC products after registration. Further, we found a non-negligible dependence of GEDITCC on the slope factor but almost independence on forest type, encouraging the spread of GEDI TCC product.
Validating GEDI tree canopy cover product across forest types using co-registered aerial LiDAR data
Gaia Vaglio Laurin;Elia Vangi;
2024
Abstract
Reliable tree canopy cover (TCC) products are vital for national forest inventory, land process modeling andforest dynamic monitoring. The new generation of space-based laser altimeter, GEDI, offers a three-dimensional(3D) insight on the forest structure, shaping the paradigm of structural variable estimation. However, the generality of newly released GEDI level-2 TCC product version 2 was less investigated across various forest types.Additionally, satellite-derived product validation usually suffers from the geolocation mismatch between satelliteand reference data. In this study, we comprehensively validated the GEDI TCC product across seven forest typesusing the reference TCC derived from several public and private aerial LiDAR datasets after geographicalregistration, and crossly compared with a commonly-used passive satellite product (i.e., GFCC TCC). As thereference aerial TCC maps were derived using various aerial LiDAR instruments, we investigated the consistencyof TCC estimation among them using simulation datasets and found that the distributions of TCC relative bias(biasR, %) were almost identical and the differences of relative RMSE (rRMSE, %) was less than 0.2%. Throughthe registration process, we found that the geolocation offsets of GEDI footprints tended to be independent ofazimuth directions and their average was about 10 m, verifying the necessity of registration during the validationprocess. Importantly, the post-registration validation of GEDI TCC showed an average RMSE of 0.10 and anaverage R2 of 0.85 for all forest types, resulting in a decrease of RMSE of up to 0.15 and an increase of R2 of up to0.33 compared to the pre-registration validation. The inter-comparison also exhibited improved consistencybetween GEDI and GFCC TCC products after registration. Further, we found a non-negligible dependence of GEDITCC on the slope factor but almost independence on forest type, encouraging the spread of GEDI TCC product.File | Dimensione | Formato | |
---|---|---|---|
VaglioLaurin_ISPRS_2024.pdf
solo utenti autorizzati
Descrizione: Validating GEDI tree canopy cover product across forest types using co-registered aerial LiDAR data
Tipologia:
Versione Editoriale (PDF)
Licenza:
NON PUBBLICO - Accesso privato/ristretto
Dimensione
8.4 MB
Formato
Adobe PDF
|
8.4 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.