Data processing for geospatial predictors

Raster and Vector data

We import and project geospatial data derived from digital elevation models (DEMs)

In particular, geomorphons is derived using Whitebox Tools which implements the algorithm from

Jasiewicz, J., and Stepinski, T. F. (2013). Geomorphons — a pattern recognition approach to classification and mapping of landforms. Geomorphology, 182, 147-156.

The 10 types of geomorphons are given categorical letter codes


catconv <- data.frame(id = 1:10, geomorphons = letters[1:10])

# Colville
col_hli <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/COL/col_HLI.tif") |>
    project("EPSG:26910")
col_geomorph <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/COL/col_geomorphons_11_23.tif")
col_geomorph_cat <- categories(col_geomorph, layer = 1, value = catconv) |>
    project("EPSG:26910")


# Mashel
mas_hli <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/MAS/mas_HLI.tif") |>
    project("EPSG:26910")
mas_geomorph <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/MAS/mas_geomorphons_11_23.tif") |>
    project("EPSG:26910")
mas_geomorph_cat <- categories(mas_geomorph, layer = 1, value = catconv) |>
    project("EPSG:26910")


# Hoh
hoh_hli <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/HOH/hoh_HLI.tif") |>
    project("EPSG:26910")
hoh_geomorph <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/HOH/hoh_geomorphons_11_23.tif")
hoh_geomorph_cat <- categories(hoh_geomorph, layer = 1, value = catconv) |>
    project("EPSG:26910")

We merge layers to make extracting geospatial predictors easier

The merge_wip refers to the Wetland Intrinsic Potential (WIP) tool and is derived from implementing the algorithm in

Halabisky, Meghan, Dan Miller, Anthony J. Stewart, Amy Yahnke, Daniel Lorigan, Tate Brasel, and Ludmila Monika Moskal. “The Wetland Intrinsic Potential Tool: Mapping Wetland Intrinsic Potential through Machine Learning of Multi-Scale Remote Sensing Proxies of Wetland Indicators.” Hydrology and Earth System Sciences 27, no. 20 (October 20, 2023): 3687–99. https://doi.org/10.5194/hess-27-3687-2023.

# Merge Layers
geomorph_stack <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/ALL_WA_geomorph_merge.tif")  #terra::merge(col_geomorph_cat, mas_geomorph_cat, hoh_geomorph_cat, filename = 'SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/ALL_WA_geomorph_merge.tif')
hli_stack <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/ALL_WA_HLI_merge.tif")  #terra::merge(col_hli, mas_hli, hoh_hli, filename = 'SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/ALL_WA_HLI_merge.tif')

merge_wip <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/ALL_WA_WIP_MERGE.tif")

Here is a map of the Hoh River Watershed WIP, where 0.0 = 0% Wetland or Upland and 1.0 = 100% Wetland

The gradient captures hydrologic conditions across the landscape and derived mostly from topography metrics at different scales

More geospatial predictors from across WA state. This includes precipitation and temperature from The PRISM Climate Group

PRISM Climate Group, Oregon State University, https://prism.oregonstate.edu

Spectral data are considered to approximate vegetation productivity and health. They are derived from Landsat median reflectance data collected over 2013-2022 and downloaded from Google Earth Engine.

Geology data on geologic age and lithology are downloaded from the WA Dept. of Natural Resoures

# All WA layers
wa_state <- vect("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/WA_State_Boundary/WA_State_Boundary.shp") |>
    project("EPSG:4326")

wa_precip <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/ALL_WA_Precip.tif") |>
    terra::project("EPSG:26910")
wa_temp <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/ALL_WA_Temp.tif") |>
    terra::project("EPSG:26910")

wa_geology <- vect("WA_Geo/ger_portal_surface_geology_100k/surface_geology_100k.gdb",
    layer = "geologic_unit_poly_100k") |>
    terra::project("EPSG:26910") |>
    tidyterra::select(LITHOLOGY, GEOLOGIC_AGE)

wa_spec1 <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/WA_median_region/WA_median_regions-0000000000-0000000000.tif")
wa_spec2 <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/WA_median_region/WA_median_regions-0000000000-0000019968.tif")  #
wa_spec3 <- rast("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/WA_median_region/WA_median_regions-0000006656-0000006656.tif")  #

wa_spec_merge <- terra::merge(wa_spec1, wa_spec2, wa_spec3, filename = "SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/WA_median_GEE_spec_median_merge.tif",
    overwrite = T) |>
    project("EPSG:26910")

Extraction

The locations of the pedon observations are used to extract all geospatial metrics.


wa_geo_ext <- read.csv("SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/WA_geo_extract.csv")

wa_dat_pts_ext <- wa_dat_pts |>
    terra::extract(x = geomorph_stack, bind = TRUE, method = "simple") |>
    terra::extract(x = hli_stack, bind = TRUE, method = "simple") |>
    terra::extract(x = wa_precip, bind = TRUE, method = "simple") |>
    terra::extract(x = wa_temp, bind = TRUE, method = "simple") |>
    terra::extract(x = merge_wip, bind = TRUE, method = "simple") |>
    terra::extract(x = wa_spec_merge, bind = TRUE, method = "simple") |>
    tidyterra::rename(HLI = lyr.1, Precip = PRISM_ppt_30yr_normal_800mM4_annual_asc,
        Temp = PRISM_tmean_30yr_normal_800mM4_annual_asc, WIP = WET) |>
    tidyterra::select(-uncertainty) |>
    tidyterra::bind_spat_cols(wa_geo_ext[, c("LITHOLOGY", "GEOLOGIC_AGE")])
(values(wa_dat_pts_ext))
names(wa_dat_pts_ext)

write.csv(wa_dat_pts_ext, file = "SOIL CARBON/ANALYSIS/R/WA_SOC_controls/All_WA_horizons_spec_geoagelith.csv")
writeVector(wa_dat_pts_ext, file = "SOIL CARBON/SPATIAL LAYERS/SPATIAL_LAYERS_7_11_22/All_WA/All_WA_horizons_spec_geoagelith.gpkg",
    overwrite = T)

Data processing for geospatial predictors

Anthony Stewart

Point Data

Raster and Vector data

Extraction