| Title: | Land Productivity Dynamics Indicator |
|---|---|
| Description: | It uses 'phenological' and productivity-related variables derived from time series of vegetation indexes, such as the Normalized Difference Vegetation Index, to assess ecosystem dynamics and change, which eventually might drive to land degradation. The final result of the Land Productivity Dynamics indicator is a categorical map with 5 classes of land productivity dynamics, ranging from declining to increasing productivity. See www.sciencedirect.com/science/article/pii/S1470160X21010517/ for a description of the methods used in the package to calculate the indicator. |
| Authors: | Xavier Rotllan-Puig [aut, cre], Eva Ivits [aut], Michael Cherlet [aut] |
| Maintainer: | Xavier Rotllan-Puig <[email protected]> |
| License: | GPL-3 |
| Version: | 1.0.5 |
| Built: | 2025-03-12 06:33:40 UTC |
| Source: | https://github.com/xavi-rp/lpdynr |
baseline_lev() derives land productivity at the beginning of the time series on study, resulting in a 3-class SpatRaster object with (1) low, (2) medium and (3) high productivity
baseline_lev( obj2process = NULL, yearsBaseline = 3, drylandProp = 0.4, highprodProp = 0.1, cores2use = 1, filename = "" )baseline_lev( obj2process = NULL, yearsBaseline = 3, drylandProp = 0.4, highprodProp = 0.1, cores2use = 1, filename = "" )
obj2process |
SpatRaster object (or its file name). If time series, each layer is one year |
yearsBaseline |
Numeric. Number of years to be averaged and used as baseline. Optional. Default is 3 |
drylandProp |
Numeric. Proportion of drylands over total land, either expressed as a fraction of unity or percentage. Optional. Default is 0.4 |
highprodProp |
Numeric. Proportion of land classified as 'highly productive' over total land, either expressed as a fraction of unity or percentage. Optional. Default is 0.1 |
cores2use |
Numeric. Number of cores to use for parallelization. Optional. Default is 1 (no parallelization) |
filename |
Character. Output filename. Optional |
baseline_lev() uses the proportion of drylands over the total land ('drylandProp') to classify the level of productivity into low level. UNPD declares that 40 percent of the World’s land resources are drylands (Middleton et al., 2011) and, therefore, 40 percent of pixels at the global level can be classified as low productivity land. This assumption is the default, but it should be adjusted for local and regional studies. In addition, baseline_lev() classifies by default 10 percent of pixels as high level of land productivity and the rest (100 - ('drylandProp' + 10)) as medium level. Proportion of pixels classified as 'high' can be also modified by passing the argument 'highprodProp'
SpatRaster object
Xavier Rotllan-Puig
Middleton, N., L. Stringer, A. Goudie, and D. Thomas. 2011. “The Forgotten Billion. MDG Achievement in the Drylands.” New York, NY, 10017, USA: United Nations Development Programme.
sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) baseline_lev(obj2process = sb, yearsBaseline = 3, drylandProp = 0.4, cores2use = 2)sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) baseline_lev(obj2process = sb, yearsBaseline = 3, drylandProp = 0.4, cores2use = 2)
clust_optim produces a scree plot with number of cluster at x-axis and total within-cluster sum of squares at y-axis
clust_optim( obj2clust = NULL, num_clstrs = seq(5, 50, 5), standardise_vars = TRUE, ... )clust_optim( obj2clust = NULL, num_clstrs = seq(5, 50, 5), standardise_vars = TRUE, ... )
obj2clust |
SpatRaster object (or its file name). Each layer is one variable |
num_clstrs |
Numeric. Optional. Vector with a sequence of number of clusters to check for optimal |
standardise_vars |
Logical. Optional. If TRUE (default), variables are standardised (mean = 0; sd = 1) |
... |
Optional. Arguments for |
The 'scree plot method' allows the user to assess how the quality of the
K-means clustering improves when increasing the number of clusters. An elbow in the curve
indicates the optimal number of clusters. K-means are run with kmeans
A scree plot
Xavier Rotllan-Puig
dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) clust_optim(obj2clust = variables_noCor, num_clstrs = seq(5, 50, 5))dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) clust_optim(obj2clust = variables_noCor, num_clstrs = seq(5, 50, 5))
EFT_clust derives the Ecosystem Functional Types using K-means to perform a clustering on the pixels of the SpatRaster object
EFT_clust( obj2clust = NULL, n_clust = 20, standardise_vars = TRUE, filename = "", ... )EFT_clust( obj2clust = NULL, n_clust = 20, standardise_vars = TRUE, filename = "", ... )
obj2clust |
SpatRaster object (or its file name). Each layer is one variable |
n_clust |
Numeric. Number of total clusters. Optional. Default = 20 |
standardise_vars |
Logical. Optional. If TRUE (default), variables are standardised (mean = 0; sd = 1) |
filename |
Character. Output filename. Optional |
... |
Arguments for |
kmeans does not optimize the final number of clusters. It needs to be set by means of 'n_clust'
(default = 20). There are several methods and statistics to determine the optimal number. clust_optim
produces a scree plot to help the user to decide the optimal number of clusters.
EFT_clust passes as default to kmeans iter.max = 500 and algorithm = "MacQueen", but these can be
modified passing these arguments through '...'
Please note that the variables are standardised (mean = 0; sd = 1) before running the clustering
An evaluation of the clustering is provided together with the SpatRaster object.
It is calculated as model$betweenss / model$totss * 100;
where 'betweenss' and 'totss' are generated by kmeans
A list with two components: (1) a SpatRaster object with the clusters and (2) a vector with the clustering evaluation in percentage
Xavier Rotllan-Puig
PCAs4clust; clust_optim; kmeans
dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) EFT_clust(obj2clust = variables_noCor, n_clust = 10)dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) EFT_clust(obj2clust = variables_noCor, n_clust = 10)
LNScaling (Local Net Productivity Scaling) uses a productivity variable (SpatRaster), e.g. season growth, to calculate the actual status of land productivity relative to its potential in homogeneous land areas or Ecosystem Functional Types (SpatRaster). If the productivity variable 'ProdVar' is a SpatRaster object with time series, it is calculated the average of the last 5 years
LNScaling(EFTs = NULL, ProdVar = NULL, cores2use = 1, filename = "")LNScaling(EFTs = NULL, ProdVar = NULL, cores2use = 1, filename = "")
EFTs |
SpatRaster object (or its file name). Ecosystem Functional Types. Its first variable has the number of EFT (cluster) each pixel belongs to |
ProdVar |
SpatRaster object (or its file name). Productivity variable (e.g. Cyclic fraction -season growth-) |
cores2use |
Numeric. Number of cores to use for parallelization. Optional. Default is 1 (no parallelization) |
filename |
Character. Output filename. Optional |
The Local Net Primary Production Scaling (LNS) method (Prince, 2009) calculates the difference between the potential and actual Net Primary Production for each pixel in homogeneous land areas. The current land production related to the local potential reflects the current level of productivity efficiency and, therefore, it is useful for the delineation of a land productivity status map
SpatRaster object
Xavier Rotllan-Puig
Prince, S.D., Becker-Reshef, I. and Rishmawi, K. 2009. “Detection and Mapping of Long-Term Land Degradation Using Local Net Production Scaling: Application to Zimbabwe.” REMOTE SENSING OF ENVIRONMENT 113 (5): 1046–57
dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) EFTs_raster <- EFT_clust(obj2clust = variables_noCor, n_clust = 10) sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) LNScaling(EFTs = EFTs_raster[[1]], ProdVar = sb)dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) EFTs_raster <- EFT_clust(obj2clust = variables_noCor, n_clust = 10) sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) LNScaling(EFTs = EFTs_raster[[1]], ProdVar = sb)
LongTermChange combines the Steadiness Index with the baseline levels of land productivity and with the change of state along the time series, resulting in a 22-class object (see details)
LongTermChange( SteadinessIndex = NULL, BaselineLevels = NULL, StateChange = NULL, filename = "" )LongTermChange( SteadinessIndex = NULL, BaselineLevels = NULL, StateChange = NULL, filename = "" )
SteadinessIndex |
SpatRaster object (or its file name). Steadiness Index (4-class) |
BaselineLevels |
SpatRaster object (or its file name). Baseline levels of land productivity (beginning of time series; 3-class) |
StateChange |
SpatRaster object (or its file name). Change of state of land productivity (beginning minus end of time series; 3-class) |
filename |
Character. Output filename. Optional |
St1-low-No Change <- 1
St1-low-Change 1 categ <- 2
St1-low-Change 2 or more categs <- 3
St1-medium-No Change <- 4
St1-medium-Change 1 categ <- 5
St1-medium-Change 2 or more categs <- 6
St1-high-No Change <- 7
St1-high-Change 1 categ <- 8
St1-high-Change 2 or more categs <- 9
St2-low-No Change <- 10
St2-low-Change 1 categ <- 10
St2-low-Change 2 or more categs <- 10
St2-medium-No Change <- 11
St2-medium-Change 1 categ <- 11
St2-medium-Change 2 or more categs <- 11
St2-high-No Change <- 12
St2-high-Change 1 categ <- 12
St2-high-Change 2 or more categs <- 12
St3-low-No Change <- 13
St3-low-Change 1 categ <- 13
St3-low-Change 2 or more categs <- 13
St3-medium-No Change <- 14
St3-medium-Change 1 categ <- 14
St3-medium-Change 2 or more categs <- 14
St3-high-No Change <- 15
St3-high-Change 1 categ <- 15
St3-high-Change 2 or more categs <- 15
St4-low-No Change <- 16
St4-low-Change 1 categ <- 17
St4-low-Change 2 or more categs <- 18
St4-medium-No Change <- 19
St4-medium-Change 1 categ <- 20
St4-medium-Change 2 or more categs <- 21
St4-high-No Change <- 22
St4-high-Change 1 categ <- 22
St4-high-Change 2 or more categs <- 22
Values = 0 in the final map indicates that there is a scarcity of data in the productivity variable (i.e. only 1 year with data), so that the indicator cannot be calculated
SpatRaster object
Xavier Rotllan-Puig
steadiness, baseline_lev, state_change
sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) SteadinessIndex_raster <- steadiness(obj2process = sb) BaselineLevels_raster <- baseline_lev(obj2process = sb, yearsBaseline = 3, drylandProp = 0.4) StateChange_raster <- state_change(obj2process = sb, yearsBaseline = 3) LongTermChange(SteadinessIndex = SteadinessIndex_raster, BaselineLevels = BaselineLevels_raster, StateChange = StateChange_raster)sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) SteadinessIndex_raster <- steadiness(obj2process = sb) BaselineLevels_raster <- baseline_lev(obj2process = sb, yearsBaseline = 3, drylandProp = 0.4) StateChange_raster <- state_change(obj2process = sb, yearsBaseline = 3) LongTermChange(SteadinessIndex = SteadinessIndex_raster, BaselineLevels = BaselineLevels_raster, StateChange = StateChange_raster)
LPD_CombAssess combines a 'LandProd_change' map (SpatRaster) with a 'LandProd_current' map (SpatRaster), giving a 5-classes map ranging from declining to increasing land productivity. 'LandProd_current' is reclassified into two classes: pixels with less than 'local_prod_threshold' (in percentage; 50 by default) of potential local productivity (within the EFT) and pixels with more or equal to 'local_prod_threshold'.
If 'LandProd_current' = NULL, 'LandProd_change' is directly reclassified into the same 5-classes map without using 'LandProd_current'. See the ATBD for the way pixels are reclassified.
LPD_CombAssess( LandProd_change = NULL, LandProd_current = NULL, local_prod_threshold = 50, filename = "" )LPD_CombAssess( LandProd_change = NULL, LandProd_current = NULL, local_prod_threshold = 50, filename = "" )
LandProd_change |
SpatRaster object (or its file name). Land Productivity Long Term Change Map |
LandProd_current |
SpatRaster object (or its file name). Land Productivity Current Status Map |
local_prod_threshold |
Numeric. Potential local productivity threshold (within the Ecosystem Functional Type) in percentage. Optional. Default = 50 |
filename |
Character. Output filename. Optional |
LandProd_change c(1:6, 8:9) & LandProd_current < 'local_prod_threshold' <- 1 Declining land productivity
LandProd_change c(3, 6) & LandProd_current >= 'local_prod_threshold' <- 1 Declining land productivity
LandProd_change c(7) & LandProd_current < 'local_prod_threshold' <- 2 Early signs of decline of land productivity
LandProd_change c(1:2, 4:5, 8:9) & LandProd_current >= 'local_prod_threshold' <- 2 Early signs of decline of land productivity
LandProd_change c(7) & LandProd_current >= 'local_prod_threshold' <- 3 Negative fluctuation (stable, but stressed land prod.)
LandProd_change c(10:12) <- 3 Negative fluctuation (stable, but stressed land prod.) LandProd_change c(13:15) <- 4 Positive fluctuation (stable, not stressed land prod.)
LandProd_change c(16:17, 19) & LandProd_current < 'local_prod_threshold' <- 4 Positive fluctuation (stable, not stressed land prod.)
LandProd_change c(18, 20:22) & LandProd_current < 'local_prod_threshold' <- 5 Increasing land productivity
LandProd_change c(16:22) & LandProd_current >= 'local_prod_threshold' <- 5 Increasing land productivity
Values = 0 in the final map indicates that there is a scarcity of data in the productivity variable (i.e. only 1 year with data), so that the indicator cannot be calculated
'local_prod_threshold' is the threshold used to classify 'LandProd_current' into pixels with low or high potential productivity within its Ecosystem Functional Type
SpatRaster
Xavier Rotllan-Puig
sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) SteadinessIndex_raster <- steadiness(obj2process = sb) BaselineLevels_raster <- baseline_lev(obj2process = sb, yearsBaseline = 3, drylandProp = 0.4) StateChange_raster <- state_change(obj2process = sb, yearsBaseline = 3) LandProd_change_raster <- LongTermChange(SteadinessIndex = SteadinessIndex_raster, BaselineLevels = BaselineLevels_raster, StateChange = StateChange_raster) dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) EFTs_raster <- EFT_clust(obj2clust = variables_noCor, n_clust = 10) LandProd_current_raster <- LNScaling(EFTs = EFTs_raster[[1]], ProdVar = sb) LPD_CombAssess(LandProd_change = LandProd_change_raster, LandProd_current = LandProd_current_raster)sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) SteadinessIndex_raster <- steadiness(obj2process = sb) BaselineLevels_raster <- baseline_lev(obj2process = sb, yearsBaseline = 3, drylandProp = 0.4) StateChange_raster <- state_change(obj2process = sb, yearsBaseline = 3) LandProd_change_raster <- LongTermChange(SteadinessIndex = SteadinessIndex_raster, BaselineLevels = BaselineLevels_raster, StateChange = StateChange_raster) dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) EFTs_raster <- EFT_clust(obj2clust = variables_noCor, n_clust = 10) LandProd_current_raster <- LNScaling(EFTs = EFTs_raster[[1]], ProdVar = sb) LPD_CombAssess(LandProd_change = LandProd_change_raster, LandProd_current = LandProd_current_raster)
PCAs4clust runs a two-steps process to prepare the data to be clustered
PCAs4clust(obj2process = NULL, cumul_var_threshold = 0.9, filename = "", ...)PCAs4clust(obj2process = NULL, cumul_var_threshold = 0.9, filename = "", ...)
obj2process |
SpatRaster object (or its file name). Each layer is one variable |
cumul_var_threshold |
Numeric. Optional (default = 0.9). Threshold of cumulative variance to select the number of useful PCs |
filename |
Character. Output filename. Optional |
... |
Optional. Arguments for |
Firstly, a Principal Component Analysis ('screening PCA') with all the variables in
'obj2process' is run in order to know the optimal number of variables to be used in a subsequent
PCA, as well as the most associated variable to those Principal Components (PCs). A threshold
of cumulative variance (cumul_var_threshold; default = 0.9) is needed. Secondly, a 'final PCA'
is run with the results of the 'screening PCA' (i.e. number of PC axes and their most associated
variables). PCAs4clust uses prcomp to run PCAs
SpatRaster object
Xavier Rotllan-Puig
dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) PCAs4clust(obj2process = variables_noCor, cumul_var_threshold = 0.9)dirctry <- paste0(system.file(package='LPDynR'), "/extdata") variables_noCor <- rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7) PCAs4clust(obj2process = variables_noCor, cumul_var_threshold = 0.9)
rm_multicol calculates the average of each variable and removes those variables highly correlated
rm_multicol( dir2process = NULL, yrs2use = NULL, multicol_cutoff = 0.7, cores2use = 1, filename = "", ... )rm_multicol( dir2process = NULL, yrs2use = NULL, multicol_cutoff = 0.7, cores2use = 1, filename = "", ... )
dir2process |
Character. Directory where the SpatRaster objects are stored. All the .tif files in the directory will be read in to be used |
yrs2use |
Numeric. A numeric vector with the years (layers positions) of the time series to be used (e.g. yrs2use = 2:21). Optional. Default (= NULL) uses all years |
multicol_cutoff |
Numeric. Cutoff value of (Pearson's) correlation. Optional. Default is 0.70 |
cores2use |
Numeric. Number of cores to use for parallelization. Optional. Default is 1 (no parallelization) |
filename |
Character. Output filename. Optional |
... |
Optional. Arguments for removeCollinearity() |
Firstly, after reading in all .tif files in 'dir2process', if they are multiband (time series), averages are
calculated. Secondly, it creates a SpatRaster object with those (averaged) variables which have a Pearson's
correlation coefficient below 'multicol_cutoff'. See removeCollinearity for further arguments
and functionalities
SpatRaster object
Xavier Rotllan-Puig
Leroy B, Meynard CN, Bellard C, Courchamp F (2015). “virtualspecies, an R package to generate virtual species distributions”. Ecography. doi: 10.1111/ecog.01388
dirctry <- paste0(system.file(package='LPDynR'), "/extdata") # directory with variables to process rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7, plot = TRUE)dirctry <- paste0(system.file(package='LPDynR'), "/extdata") # directory with variables to process rm_multicol(dir2process = dirctry, multicol_cutoff = 0.7, plot = TRUE)
SpatRaster object containing time series of land productivity data (400 cells; 10 layers). Years 2000-2009. The variable has been derived from MODIS using Timesat
sb_catsb_cat
GeoTIFF. SpatRaster object with dimensions: 20, 20, 400, 10 (nrow, ncol, ncell, nlayers)
Standing biomass
Downloaded from www.eea.europa.eu (20/08/2020)
Coord. ref. : +proj=laea +lat_0=52 +lon_0=10 +x_0=4321000 +y_0=3210000 +ellps=GRS80 +units=m +no_defs
Resolution : 500, 500 (x, y)
Extent : 3640000, 3650000, 2140000, 2150000 (xmin, xmax, ymin, ymax)
https://www.eea.europa.eu/en/analysis
https://www.eea.europa.eu/data-and-maps/data/annual-above-ground-vegetation-productivity
terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif"))terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif"))
SpatRaster object containing time series of phenological data (400 cells; 10 layers). Years 2000-2009. The variable has been derived from MODIS using Timesat
sbd_catsbd_cat
GeoTIFF. SpatRaster object with dimensions: 20, 20, 400, 10 (nrow, ncol, ncell, nlayers)
Season beginning day
Downloaded from www.eea.europa.eu (20/08/2020)
Coord. ref. : +proj=laea +lat_0=52 +lon_0=10 +x_0=4321000 +y_0=3210000 +ellps=GRS80 +units=m +no_defs
Resolution : 500, 500 (x, y)
Extent : 3640000, 3650000, 2140000, 2150000 (xmin, xmax, ymin, ymax)
https://www.eea.europa.eu/en/analysis
https://www.eea.europa.eu/data-and-maps/data/annual-start-of-vegetation-growing
terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sbd_cat.tif"))terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sbd_cat.tif"))
SpatRaster object containing time series of phenological data (400 cells; 10 layers). Years 2000-2009. The variable has been derived from MODIS using Timesat
sl_catsl_cat
GeoTIFF. SpatRaster object with dimensions: 20, 20, 400, 10 (nrow, ncol, ncell, nlayers)
Season Length
Downloaded from www.eea.europa.eu (20/08/2020)
Coord. ref. : +proj=laea +lat_0=52 +lon_0=10 +x_0=4321000 +y_0=3210000 +ellps=GRS80 +units=m +no_defs
Resolution : 500, 500 (x, y)
Extent : 3640000, 3650000, 2140000, 2150000 (xmin, xmax, ymin, ymax)
https://www.eea.europa.eu/en/analysis
https://www.eea.europa.eu/data-and-maps/data/annual-above-ground-vegetation-season
terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sl_cat.tif"))terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sl_cat.tif"))
state_change derives land productivity state change between the beginning and the end of the time series on study, resulting in a 3-class SpatRaster object with (1) no change, (2) changed between 1 and x classes or (3) changed more than x classes, where x can be defined by the user (default is 1)
state_change( obj2process = NULL, yearsBaseline = 3, changeNclass = 1, cores2use = 1, filename = "" )state_change( obj2process = NULL, yearsBaseline = 3, changeNclass = 1, cores2use = 1, filename = "" )
obj2process |
SpatRaster object (or its file name). If time series, each layer is one year |
yearsBaseline |
Numeric. Number of years to be averaged at the beginning and end of the time series. Optional. Default is 3 |
changeNclass |
Numeric. Number of classes changed for classification. Optional. Default is 1 |
cores2use |
Numeric. Number of cores to use for parallelization. Optional. Default is 1 (no parallelization) |
filename |
Character. Output filename. Optional |
state_change uses the average of 'yearsBaseline' number of years at the beginning and the end of the time series
SpatRaster object
Xavier Rotllan-Puig
sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) state_change(obj2process = sb, yearsBaseline = 3, changeNclass = 1, cores2use = 2)sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) state_change(obj2process = sb, yearsBaseline = 3, changeNclass = 1, cores2use = 2)
steadiness derives the Steadiness Index from a land productivity variable
steadiness(obj2process = NULL, cores2use = 1, filename = "")steadiness(obj2process = NULL, cores2use = 1, filename = "")
obj2process |
SpatRaster object (or its file name). If time series, each layer is one year |
cores2use |
Numeric. Number of cores to use for parallelization. Optional. Default is 1 (no parallelization) |
filename |
Character. Output filename. Optional |
The Steadiness Index is based on the combination of two metrics calculated per pixel: (1) the slope derived from a linear regression of the different years of the time series and (2) the net change on the same period. It results in a 4-class SpatRaster object ranging from (1) strong negative to (4) strong positive ecosystem dynamics. See Ivits et al. (2013) for further explanations.
Values = 0 in the final map indicates that there is a scarcity of data in the productivity variable (i.e. only 1 year with data), so that the indicator cannot be calculated
SpatRaster object
Xavier Rotllan-Puig
Ivits, E., M. Cherlet, W. Mehl, and S. Sommer. 2013. “Ecosystem Functional Units Characterized by Satellite Observed Phenology and Productivity Gradients: A Case Study for Europe.” Ecological Indicators 27: 17–28. doi:10.1016/j.ecolind.2012.11.010
sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) steadiness(obj2process = sb)sb <- terra::rast(paste0(system.file(package='LPDynR'), "/extdata/sb_cat.tif")) steadiness(obj2process = sb)