Lidar data

[Friessn]

We need lidar indices for forest regions of Rheinland-Pfalz using the same raster as the sentinel data.
Excluding vegetation < 10m was not straightforward in the RSDB, @swoellauer will help us extracting the indices.

[Friessn]

We calculate the following indices:
BE_ELEV_MEAN
BE_ELEV_SLOPE
surface ratio (of DTM,,DSM,CHM)
TCH
vergetation coverage (1,2,5,10m)
BE_H_MEDIAN
BE_H_KURTOSIS
BE_H_MAX
BE_H_SD
BE_H_SKEW
BE_H_P30, BE_H_P70
LAI
BE_PR_CAN
BE_PR_UND
BE_PR_REG
BE_RD_CAN
BE_RD_UND
BE_RD_REG
BE_FHD
surface intensity mean
ABG
VDR
chm_height_max
chm_height_mean
chm_height_sd
chm_survace_ratio
dtm_elevation_max
dtm_elevation_sd

[swoellauer]

Task is running with indices:

BE_ELEV_MEAN, BE_ELEV_SLOPE, dtm_surface_ratio, dsm_surface_ratio, chm_surface_ratio, TCH, vegetation_coverage_01m_CHM, vegetation_coverage_02m_CHM, vegetation_coverage_05m_CHM, vegetation_coverage_10m_CHM, BE_H_MEDIAN, BE_H_KURTOSIS, BE_H_MAX, BE_H_SD, BE_H_SKEW, BE_H_P30, BE_H_P70, LAI, BE_PR_CAN, BE_PR_UND, BE_PR_REG, BE_RD_CAN, BE_RD_UND, BE_RD_REG, BE_FHD, surface_intensity_mean, AGB, VDR, chm_height_max, chm_height_mean, chm_height_sd, dtm_elevation_max, dtm_elevation_sd

Task:

{ "task_pointcloud": "index_raster", "pointcloud": "rheinland_pfalz", "rasterdb": "RLP_forest_mask_20m_i4", "indices": [ "BE_ELEV_MEAN", "BE_ELEV_SLOPE", "dtm_surface_ratio", "dsm_surface_ratio", "chm_surface_ratio", "TCH", "vegetation_coverage_01m_CHM", "vegetation_coverage_02m_CHM", "vegetation_coverage_05m_CHM", "vegetation_coverage_10m_CHM", "BE_H_MEDIAN", "BE_H_KURTOSIS", "BE_H_MAX", "BE_H_SD", "BE_H_SKEW", "BE_H_P30", "BE_H_P70", "LAI", "BE_PR_CAN", "BE_PR_UND", "BE_PR_REG", "BE_RD_CAN", "BE_RD_UND", "BE_RD_REG", "BE_FHD", "surface_intensity_mean", "AGB", "VDR", "chm_height_max", "chm_height_mean", "chm_height_sd", "dtm_elevation_max", "dtm_elevation_sd" ], "mask_band": 1 } 

[aliceziegler]

additionally:
pulse_return_mean
BE_H_10

@swoellauer: something like openness/skyview factor/gapfraction would be nice. I used gap_fraction for the Kili project. Did I just miss it in the database indices? Perhaps I calculated it seperately from the points directly.

[swoellauer]

Processing Done. 121448.323s -> 34 hours
Raster layer: RLP_forest_mask_20m_i4

More indices can be added with a new processing task.
Currently indices suggested from @aliceziegler are not included in the raster.

[Friessn]

Raw Lidar Data from the flight in 2013 is missing although DEM data at the respective locations is available. The area that is missing coincides with the border from the UTM Zones 31 and 32, maybe this is a clue.
@JannisGottwald may check with the data provider whether something went missing on their side
@swoellauer continues having his eyes open.

The missing area is quite important as it includes the Eifel. Nevertheless, we will proceed with the current plan.

[Friessn]

additionally:
pulse_return_mean
BE_H_10

@swoellauer: something like openness/skyview factor/gapfraction would be nice. I used gap_fraction for the Kili project. Did I just miss it in the database indices? Perhaps I calculated it seperately from the points directly.

I would also add the topographic wetness index (TWI) to the list. @swoellauer the TWI is currently not implemented in the rsdb, is it possible to add it?

[swoellauer]

TWI depends on flow pathways from DTM. As the flow accumulates over possibly large areas it does not fit well in the approach of calculating indices from small local areas on-demand.