movecost: R function for calculating accumulated anisotropic...

In movecost: Calculation of Slope-Dependant Accumulated Cost Surface, Least-Cost Paths, Least-Cost Corridors, Least-Cost Networks Related to Human Movement Across the Landscape

R function for calculating accumulated anisotropic slope-dependant cost of movement across the terrain and least-cost paths from a point origin

Description

The function provides the facility to calculate the anisotropic accumulated cost of movement around a starting location and to optionally calculate least-cost path(s) toward one or multiple destinations. It implements different cost estimations related to human movement across the landscape. The function takes as input a Digital Terrain Model ('RasterLayer' class) and a point feature ('SpatialPointsDataFrame' class), the latter representing the starting location, i.e. the location from which the accumulated cost is calculated. Besides citing this package, you may want to refer to the following journal article, where an earlier version of the package is described: Alberti (2019) <doi:10.1016/j.softx.2019.100331>.
Visit this LINK to access the package's vignette.

Usage

movecost(
  dtm = NULL,
  origin,
  destin = NULL,
  studyplot = NULL,
  barrier = NULL,
  plot.barrier = FALSE,
  irregular.dtm = FALSE,
  funct = "t",
  time = "h",
  outp = "r",
  move = 16,
  field = 0,
  cogn.slp = FALSE,
  sl.crit = 10,
  W = 70,
  L = 0,
  N = 1,
  V = 1.2,
  z = 9,
  return.base = FALSE,
  rb.lty = 2,
  breaks = NULL,
  cont.lab = TRUE,
  destin.lab = TRUE,
  cex.breaks = 0.6,
  cex.lcp.lab = 0.6,
  graph.out = TRUE,
  transp = 0.5,
  oneplot = TRUE,
  export = FALSE
)

Arguments

dtm	Digital Terrain Model (RasterLayer class); if not provided, elevation data will be acquired online for the area enclosed by the 'studyplot' parameter (see Details).
origin	location from which the cost surface is calculated (SpatialPointsDataFrame class).
destin	location(s) to which least-cost path(s) is calculated (SpatialPointsDataFrame class).
studyplot	polygon (SpatialPolygonDataFrame class) representing the study area for which online elevation data are acquired (see Details); NULL is default.
barrier	area where the movement is inhibited (SpatialLineDataFrame or SpatialPolygonDataFrame class).
plot.barrier	TRUE or FALSE (default) if the user wants or does not want the barrier to be plotted (in blue).
irregular.dtm	TRUE or FALSE (default) if the input DTM features irregular margins (Details).
funct	cost function to be used: -functions expressing cost as walking time- t (default) uses the on-path Tobler's hiking function; tofp uses the off-path Tobler's hiking function; mp uses the Marquez-Perez et al.'s modified Tobler's function; icmonp uses the Irmischer-Clarke's hiking function (male, on-path); icmoffp uses the Irmischer-Clarke's hiking function (male, off-path); icfonp uses the Irmischer-Clarke's hiking function (female, on-path); icfoffp uses the Irmischer-Clarke's hiking function (female, off-path); ug uses the Uriarte Gonzalez's walking-time cost function; ma uses the Marin Arroyo's walking-time cost function; alb uses the Alberti's Tobler hiking function modified for pastoral foraging excursions; gkrs uses the Garmy, Kaddouri, Rozenblat, and Schneider's hiking function; r uses the Rees' hiking function; ks uses the Kondo-Seino's hiking function; trp uses the Tripcevich's hiking function; -functions for wheeled-vehicles- wcs uses the wheeled-vehicle critical slope cost function; -functions expressing abstract cost- ree uses the relative energetic expenditure cost function; b uses the Bellavia's cost function; e uses the Eastman's cost function; -functions expressing cost as metabolic energy expenditure- p uses the Pandolf et al.'s metabolic energy expenditure cost function; pcf uses the Pandolf et al.'s cost function with correction factor for downhill movements; m uses the Minetti et al.'s metabolic energy expenditure cost function; hrz uses the Herzog's metabolic energy expenditure cost function; vl uses the Van Leusen's metabolic energy expenditure cost function; ls uses the Llobera-Sluckin's metabolic energy expenditure cost function; a uses the Ardigo et al.'s metabolic energy expenditure cost function; h uses the Hare's metabolic energy expenditure cost function (for all the mentioned cost functions, see Details).
time	time-unit expressed by the accumulated raster and by the isolines if Tobler's and other time-related cost functions are used; 'h' for hour, 'm' for minutes.
outp	type of output: 'raster' or 'contours' (see Details).
move	number of directions in which cells are connected: 4 (rook's case), 8 (queen's case), 16 (knight and one-cell queen moves; default).
field	value assigned to the cells coinciding with the barrier (0 by default).
cogn.slp	TRUE or FALSE (default) if the user wants or does not want the 'cognitive slope' to be used in place of the real slope (see Details).
sl.crit	critical slope (in percent), typically in the range 8-16 (10 by default) (used by the wheeled-vehicle cost function; see Details).
W	walker's body weight (in Kg; 70 by default; used by the Pandolf's and Van Leusen's cost function; see Details).
L	carried load weight (in Kg; 0 by default; used by the Pandolf's and Van Leusen's cost function; see Details).
N	coefficient representing ease of movement (1 by default) (see Details).
V	speed in m/s (1.2 by default) (used by the Pandolf et al.'s, Pandolf et al.s with correction factor, Van Leusen's, and Ardigo et al.'s cost function; if set to 0, it is internally worked out on the basis of Tobler on-path hiking function; see Details).
z	zoom level for the elevation data downloaded from online sources (0 to 15; 9 by default) (see Details and get_elev_raster).
return.base	TRUE or FALSE (default) if the user wants or does not want the least-cost paths back to the origin to be calculated and plotted (as dashed lines).
rb.lty	line type used to represent the least-cost paths back to the origin in the returned plot (2 by default; dashed line; see 'lty' parameter in par).
breaks	contour interval; if no value is supplied, the interval is set by default to 1/10 of the range of values of the accumulated cost surface.
cont.lab	if set to TRUE (default) display the labels of the contours over the accumulated cost surface.
destin.lab	if set to TRUE (default) display the label(s) indicating the cost at the destination location(s).
cex.breaks	set the size of the cost labels used in the contour plot (0.6 by default).
cex.lcp.lab	set the size of the labels used in least-cost path(s) plot (0.6 by default).
graph.out	TRUE (default) or FALSE if the user wants or does not want a graphical output to be generated.
transp	set the transparency of the slopeshade raster that is plotted over the rendered plots (0.5 by default).
oneplot	TRUE (default) or FALSE if the user wants or does not want the plots displayed in a single window.
export	TRUE or FALSE (default) if the user wants or does not want the outputs to be exported; if TRUE, the DTM, the cost-surface, and the accumulated cost surface are exported as a GeoTiff file, while the isolines, the least-cost path(s), and a copy of the input destination locations (storing the cost measured at each location) are exported as shapefile; all the exported files (excluding the DTM) will bear a suffix corresponding to the cost function selected by the user. Note that the DTM is exported only if it was not provided by the user and downloaded by the function from online sources.

Details

If the parameter destin is fed with a dataset representing destination location(s) ('SpatialPointsDataFrame' class), the function also calculates least-cost path(s) plotted on the input DTM; the length of each path is saved under the variable 'length' stored in the 'LCPs' dataset ('SpatialLines' class) returned by the function. In the produced plot, the red dot(s) representing the destination location(s) are labelled with numeric values representing the cost value at the location(s).

The cost value is also appended to the updated destination locations dataset returned by the function, which stores a new variable labelled cost. If the cost is expressed in terms of walking time, the labels accompaining each destinaton location will express time in sexagesimal numbers (hours, minutes, seconds). In this case, the variable 'cost' appended to the returned destination location datset will store the time figures in decimal numbers, while another variable named cost_hms will store the corresponding value in sexagesimal numbers. When interpreting the time values stored in the cost variable, the user may want to bear in mind the selected time unit (see right below).

When using cost functions expressing cost in terms of time, the time unit can be selected by the user setting the time parameter to h (hours) or to m (minutes).

In general, the user can also select which type of visualization the function has to produce; this is achieved setting the outp parameter to either r (=raster) or to c (=contours). The former will produce a raster with a colour scale and contour lines representing the accumulated cost surface; the latter parameter will only produce contour lines.

The contour lines' interval is set using the breaks parameter; if no value is passed to the parameter, the interval will be set by default to 1/10 of the range of values of the accumulated cost surface.

It is worth reminding the user(s) that all the input layers (i.e., DTM, start location, and destination locations) must use the same projected coordinate system.

Cost surface calculation:
for the cost-surface and LCPs calculation, movecost() builds on functions from Jacob van Etten's gdistance package. Under the hood, movecost() calculates the slope as rise over run, following the procedure described by van Etten, "R Package gdistance: Distances and Routes on Geographical Grids" in Journal of Statistical Software 76(13), 2017, pp. 14-15. The number of directions in which cells are connected in the cost calculation can be set to 4 (rook's case), 8 (queen's case), or 16 (knight and one-cell queen moves) using the move parameter (see 'Arguments').

Inhibition of movement (barrier):
areas where the movement is inhibited can be fed into the analysis via the barrier parameter; SpatialLineDataFrame or SpatialPolygonDataFrame can be used. The barrier is assigned a conductance value of 0 (i.e., movement is inhibited) by default, but the user can assign any other value via the field parameter. Internally, the barrier creation rests on the internal create_barrier_cs function, which relies on the same function out of an earlier version of the leastcostpath package.

To test this facility, consider the following example:

First, we use in-built data to come up with a linear feature (i.e., a LCP) that we will later use as barrier:

result1 <- movecost(volc, destin.loc[1,], destin.loc[4,])

After, we calculate the LCP between two other locations, first not using any barrier (result2), then using the mentioned LCP (from result1) as a barrier (result3):

result2 <- movecost(volc, destin.loc[3,], destin.loc[6,], move=8)
result3 <- movecost(volc, destin.loc[3,], destin.loc[6,], barrier=result1$LCPs, plot.barrier=TRUE, move=8)

As apparent by comparing result2 to result3, when the barrier is used (result3), the LCP does not cross the barrier but is "forced" to make a long detour. In result3, the barrier is plotted as a blue line. Note that the move parameter has been set to 8; if set to 16, the LCP will be "able" to jump the barrier.

DTM featuring irregular margins:
if the input DTM features irregular margins, e.g a coastline with gulfs and/or inlets where cells corresponding to the sea are given NoData, the user is to set the irregular.dtm parameter to TRUE; this will prevent the LCPs to cross the sea. Internally, what movecost() does is to generate a polygon vector layer from the DTM and to use the polygon as a mask to create a Transitional Layer via the internal create_barrier_cs function, which relies on the same function out of an earlier version from the leastcostpath package. In the mask Transitional Layer those parts corresponding to the terrain are given a conductance value equal to 1, while everything else (i.e., the parts corresponding to the sea) are given 0 conductance. The mask Transitional Layer is then internally multiplied by the conductance transitional layer representing the cost of movement (according to the user-selected function). This will set to 0 the conductance values of those parts of the study area that do not correspond to the terrain, while keeping unaltered the conductance of those parts that do coincide with the terrain.

As a case in point, let's consider the two following examples (using some in-build datasets):

resultA <- movecost(malta_dtm_40, origin=springs[5,], destin=springs[15,], irregular.dtm=FALSE, oneplot=FALSE)

resultB <- movecost(malta_dtm_40, origin=springs[5,], destin=springs[15,], irregular.dtm=TRUE, oneplot=FALSE)

As you can see, in the first case, the LCP between the two locations cross the sea, while in the second case the LCP follows the coastline. One can also appreciate the difference between the two returned conductance transitional layers:

raster::plot(raster::raster(resultA$conductance))

raster::plot(raster::raster(resultB$conductance))

It is apparent that in the second layer the sea area has been given 0 conductance, while keeping the rest unchanged. If the input DTM does not feature irregular margins (like, for instance, the built-in volc DTM), the user may safely leave the irregular.dtm parameter set to FALSE (which is the default value).

Acquiring online elevation data:
if a DTM is not provided,movecost()' will download elevation data from online sources. Elevation data will be acquired for the area enclosed by the polygon supplied by the studyplot parameter (SpatialPolygonDataFrame class). To tap online elevation data, movecost()' internally builds on the get_elev_raster function from the elevatr package.

The zoom level of the downloaded DTM (i.e., its resolution) is controlled by the parameter z, which is set to 9 by default (a trade off between resolution and download time).

To test this facility, the user may want to try the following code, that will generate a least-cost surface and least-cost paths in an area close the Mount Etna (Sicily, Italy), whose elevation data are acquired online; the start and end locations, and the polygon defining the study area, are provided in this same package:

result <- movecost(origin=Etna_start_location, destin=Etna_end_location, studyplot=Etna_boundary)

The LCPs back to the origin can be calculated and plotted setting the parameter 'return.base' to TRUE:

result <- movecost(origin=Etna_start_location, destin=Etna_end_location, studyplot=Etna_boundary, return.base=TRUE)

To know more about what elevation data are tapped from online sources, visit: https://cran.r-project.org/web/packages/elevatr/vignettes/introduction_to_elevatr.html.

For more information about the elevation data resolution per zoom level, visit https://github.com/tilezen/joerd/blob/master/docs/data-sources.md#what-is-the-ground-resolution.

To know what is sourced at what zoom level, visit https://github.com/tilezen/joerd/blob/master/docs/data-sources.md#what-is-sourced-at-what-zooms.

Terrain slope and cognitive slope:
when it comes to the terrain slope, the function provides the facility to use the so-called 'cognitive slope', following Pingel TJ (2013), Modeling Slope as a Contributor to Route Selection in Mountainous Areas, in Cartography and Geographic Information Science, 37(2), 137-148. According to Pingel, "Humans tend to overestimate geographic slopes by a surprisingly high margin...This analysis indicates downhill slopes are overestimated at approximately 2.3 times the vertical, while uphill slopes are overestimated at 2 times the vertical.". As a result, if the parameter cogn.slp is set to TRUE, positive slope values are preliminarily multiplied by 1.99, while negative slope values are multiplied by 2.31.

Terrain factor (N):
virtually all the implemented cost functions (with few exceptions) can take into account a 'terrain factor' (N parameter; 1 by default), which represents the easiness/difficulty of moving on different terrain types. According to the type of terrain, the energy spent when walking increases. The same holds true for time, which increases because on a loose terrain (for instance) the walking speed decreases. While a terrain factor is 'natively' part of the Van Leusen's, Pandolf et al.'s, and Bellavia's cost function, it has been integrated into the other cost functions as well (when/if relevant).

Note that the terrain factor has NOT been implemented in the Alberti's, Tobler's off-path, and Irmischer-Clarke's off-path cost function. As for the latter two, they already natively feature a terrain factor. Therefore, it has been implemented only in their on-path version. Needless to say, if we use a terrain factor of 1.67 with the Tobler's (on-path) hiking function, the results will be equal to those obtained using the Tobler's off-path function (the reciprocal of 1.67, i.e. 0.60, is in fact natively used by the Tobler's function for off-path hiking). In fact, compare the results of the following two runs of movecost() (using in-built datasets):

result1 <- movecost(dtm=volc, origin=volc.loc, destin=destin.loc, breaks=0.05, funct="t", N=1.67)
result2 <- movecost(dtm=volc, origin=volc.loc, destin=destin.loc, breaks=0.05, funct="tofp")

The user may want to refer to the following list of terrain factors, which is based on the data collected in Herzog, I. (2020). Spatial Analysis Based on Cost Functions. In Gillings M, Haciguzeller P, Lock G (eds), "Archaeological Spatial Analysis. A Methodological Guide.", Routledge: New York, 340 (with previous references). The list is divided into two sections (a and b), the first reporting the terrain factors to be used for cost functions measuring time, the second for functions measuring cost other than time:

(a)

• Blacktop roads, improved dirt paths, cement = 1.00

• Lawn grass = 1.03

• Loose beach sand = 1.19

• Disturbed ground (former stone quarry) = 1.24

• Horse riding path, flat trails and meadows = 1.25

• Tall grassland (with thistle and nettles) = 1.35

• Open space above the treeline (i.e., 2000 m asl) = 1.50

• Bad trails, stony outcrops and river beds = 1.67

• Off-paths = 1.67

• Bog = 1.79

• Off-path areas below the treeline (pastures, forests, heathland) = 2.00

• Rock = 2.50

• Swamp, water course = 5.00

(b)

• Asphalt/blacktop = 1.00

• Dirt road or grass = 1.10

• Hard-surface road = 1.20

• Light brush = 1.20

• Ploughed field = 1.30 or 1.50

• Heavy brush = 1.50

• Hard-packed snow = 1.60

• Swampy bog = 1.80

• Sand dunes = 1.80

• Loose sand = 2.10

Implemented cost functions:
note that in what follows x[adj] stands for slope as rise/run calculated for adjacent cells:

Tobler's hiking function (on-path) (speed in kmh):

(6 * exp(-3.5 * abs(x[adj] + 0.05))) * (1/N)

Tobler's hiking function (off-path) (speed in kmh):

(6 * exp(-3.5 * abs(x[adj] + 0.05))) * 0.6

as per Tobler's indication, the off-path walking speed is reduced by 0.6.

Marquez-Perez et al.'s modified Tobler hiking function (speed in kmh):

(4.8 * exp(-5.3 * abs((x[adj] * 0.7) + 0.03))) * (1/N)

modified version of the Tobler's hiking function as proposed by Joaquin Marquez-Parez, Ismael Vallejo-Villalta & Jose I. Alvarez-Francoso (2017), "Estimated travel time for walking trails in natural areas", Geografisk Tidsskrift-Danish Journal of Geography, 117:1, 53-62, DOI: 10.1080/00167223.2017.1316212.

Irmischer-Clarke's modified Tobler hiking function (male, on-path; speed in kmh):

((0.11 + exp(-(abs(x[adj])*100 + 5)^2 / (2 * 30^2))) * 3.6) * (1/N)

modified version of the Tobler's function as proposed for (male) on-path hiking by Irmischer, I. J., & Clarke, K. C. (2018). Measuring and modeling the speed of human navigation. Cartography and Geographic Information Science, 45(2), 177-186. https://doi.org/10.1080/15230406.2017.1292150. It is interesting to note that the hiking speed predicted by this and by the other functions proposed by the authors is slower than the one modelled by Tobler's hiking function. This is attributed to the cognition involved in wayfinding , such as map reading, analyzing the terrain, decision making, determining routes, etc. Note: all the all the Irmischer-Clarke's functions originally express speed in m/s; they have been reshaped (multiplied by 3.6) to turn m/s into km/h for consistency with the other Tobler-related cost functions; slope is in percent.

Irmischer-Clarke's modified Tobler hiking function (male, off-path; speed in kmh):

(0.11 + 0.67 * exp(-(abs(x[adj])*100 + 2)^2 / (2 * 30)^2)) * 3.6

Irmischer-Clarke's modified Tobler hiking function (female, on-path; speed in kmh):

((0.95 * (0.11 + exp(-(abs(x[adj]) * 100 + 5)^2/(2 * 30^2)))) * 3.6) * (1/N)

Irmischer-Clarke's modified Tobler hiking function (female, off-path; speed in kmh):

(0.95 * (0.11 + 0.67 * exp(-(abs(x[adj]) * 100 + 2)^2/(2 * 30^2)))) * 3.6

Uriarte Gonzalez's walking-time cost function:

1 / ((0.0277 * (abs(x[adj])*100) + 0.6115) * N)

proposed by Uriarte Gonzalez; see: Chapa Brunet, T., Garcia, J., Mayoral Herrera, V., & Uriarte Gonzalez, A. (2008). GIS landscape models for the study of preindustrial settlement patterns in Mediterranean areas. In Geoinformation Technologies for Geo-Cultural Landscapes (pp. 255-273). CRC Press. https://doi.org/10.1201/9780203881613.ch12.
The cost function originally expresses walking time in seconds; for the purpose of its implementation in this function, it is the reciprocal of time (1/T) that is used in order to eventually get T/1. Unlike in the original cost function, here the pixel resolution is not taken into account since 'gdistance' takes care of the cells' dimension when calculating accumulated costs.

Marin Arroyo's walking-time cost function:

ifelse((abs(x[adj])*100) < 0, 1 / ((0.6 * ((abs(x[adj])*100)/23+1))*N), 1 / ((0.6 * ((abs(x[adj])*100)/11+1))*N))

used by Marin Arroyo A.B. (2009), The use of optimal foraging theory to estimate Late Glacial site catchments areas from a central place: the case of eastern Cantabria, Span, in Journal of Anthropological Archaeology 28, 27-36. The cost function originally expresses walking time in seconds; here it is the reciprocal of time (1/T) that is used in order to eventually get T/1. Slope is in percent. Note: unlike in the original equation, here d (distance travelled in meter) is not taken into account since 'gdistance' takes care of the cells' dimension when calculating accumulated costs.

Alberti's Tobler hiking function modified for pastoral foraging excursions (speed in kmh):

(6 * exp(-3.5 * abs(x[adj] + 0.05))) * 0.25

proposed by Gianmarco Alberti; see: Locating potential pastoral foraging routes in Malta through the use of a Geographic Information System. The Tobler's function has been rescaled to fit animal walking speed during foraging excursions. The distribution of the latter, as empirical data show, turns out to be right-skewed and to vary along a continuum. It ranges from very low speed values (corresponding to slow grazing activities grazing while walking) to comparatively higher values (up to about 4.0 km/h) corresponding to travels without grazing (directional travel toward feeding stations). The function consider 1.5 km/h as the average flock speed, which roughly corresponds to the average speed recorded in some studies, and can be considered the typical speed of flocks during excursions in which grazing takes place while walking (typical form of grazing in most situations). Tobler's hiking function has been rescaled by a factor of 0.25 to represent the walking pace of a flock instead of humans. The factor corresponds to the ratio between the flock average speed (1.5 km/h) and the maximum human walking speed (about 6.0 km/h) on a favourable slope.

Garmy, Kaddouri, Rozenblat, and Schneider's hiking function (speed in kmh):

(4 * exp(-0.008 * ((atan(abs(x[adj]))*180/pi)^2))) * (1/N)

slope in degrees; see: Herzog, I. (2020). Spatial Analysis Based on Cost Functions. In Gillings M, Haciguzeller P, Lock G (eds), "Archaeological Spatial Analysis. A Methodological Guide.", Routledge: New York, 333-358 (with previous references).

Rees' hiking function (speed in kmh):

((1 / (0.75 + 0.09 * abs(x[adj]) + 14.6 * (abs(x[adj]))^2)) * 3.6) * (1/N)

Rees' slope-dependant cost function; it is originally expressed in terms of time (1/v in Rees' publication); here it is the reciprocal of time (i.e. speed) that is used in order to eventually get the reciprocal of speed (i.e. time). Slope is dealt with here as originally expressed in Rees' publication (i.e. rise over run). The speed, which is originally expressed in m/s, has been here transposed to kmh (i.e., multiplied by 3.6) for consistency with other hiking functions.
For this cost function see: Rees, WG (2004). Least-cost paths in mountainous terrain. Computers & Geosciences, 30(3), 203-209. See also: Campbell MJ, Dennison PE, Butler BW, Page WG (2019). Using crowdsourced fitness tracker data to model the relationship between slope and travel rates. Applied Geography 106, 93-107 (with previous references).

Kondo-Seino's modified Tobler hiking function (speed in kmh):

ifelse(abs(x[adj]) >= -0.07, (5.1 * exp(-2.25 * abs(x[adj] + 0.07))) * (1/N), (5.1 * exp(-1.5 * abs(x[adj] + 0.07)))) * (1/N)

Kondo-Seino's modified Tobler hiking function; it expresses walking speed in Kmh; slope as rise/run; see Kondo Y., Seino Y. (2010). GPS-aided Walking Experiments and Data-driven Travel Cost Modelingon the Historical Road of Nakasendo-Kisoji (Central Highland Japan), in: Frischer B., Webb Crawford J., Koller D. (eds.), Making History Interactive. Computer Applications and Quantitative Methods in Archaeology (CAA). Proceedings of the 37th International Conference, Williamsburg, Virginia, United States of America, March 22-26 (BAR International Series S2079). Archaeopress, Oxford, 158-165.

Tripcevich hiking function (speed in kmh):

((4.028*46^2)/(((atan(abs(x[adj]))*180/pi)+4.127)^2+46^2))*(1/N)

Tripcevich's hiking function; it expresses walking speed in Kmh; slope is originally expressed in degrees; see Tripcevich N (2008). Estimating Llama caravan travel speeds: ethno-archaeological fieldwork with a Peruvian salt caravan. Trabajo presentado el la inauguracionn del Centre for Spatial Studies, University of California, Santa Barbara. See also: Lucero G, Marsh EJ, Castro S (2014), Rutas prehistoricas en lo NO de San Juan: una propuesta macroregional desde los sistemas de information geografica, in Cortegoso V, Duran V, Gasco Alejandra (eds), Arqueologia de ambientes de altura de Mendoza y San Juan (Argentina), EDIUNC, pp. 275-305.

Wheeled-vehicle critical slope cost function:

1 / ((1 + ((abs(x[adj])*100) / sl.crit)^2) * N)

where sl.crit (=critical slope, in percent) is "the transition where switchbacks become more effective than direct uphill or downhill paths" and typically is in the range 8-16; see Herzog, I. (2016). Potential and Limits of Optimal Path Analysis. In A. Bevan & M. Lake (Eds.), Computational Approaches to Archaeological Spaces (pp. 179-211). New York: Routledge.

Relative energetic expenditure cost function:

1 / ((tan((atan(abs(x[adj]))*180/pi)*pi/180) / tan (1*pi/180)) * N)

slope-based cost function expressing change in potential energy expenditure; see Conolly, J., & Lake, M. (2006). Geographic Information Systems in Archaeology. Cambridge: Cambridge University Press, p. 220; see also Newhard, J. M. L., Levine, N. S., & Phebus, A. D. (2014). The development of integrated terrestrial and marine pathways in the Argo-Saronic region, Greece. Cartography and Geographic Information Science, 41(4), 379-390, with references to studies that use this function; see also ten Bruggencate, R. E., Stup, J. P., Milne, S. B., Stenton, D. R., Park, R. W., & Fayek, M. (2016). A human-centered GIS approach to modeling mobility on southern Baffin Island, Nunavut, Canada. Journal of Field Archaeology, 41(6), 684-698. https://doi.org/10.1080/00934690.2016.1234897.

Bellavia's cost function:

1 / (N * ((atan(abs(x[adj]))*180/pi)+1))

proposed by G. Bellavia, it measures abstract cost. Slope in degrees; N is a terrain factor (see above). See: Herzog I. (2020). Spatial Analysis Based on Cost Functions. In Gillings M, Haciguzeller P, Lock G (eds), "Archaeological Spatial Analysis. A Methodological Guide.", Routledge: New York, 333-358 (with previous references).

Eastman's cost function:

1 / ((0.031*(atan(abs(x[adj]))*180/pi)^2-0.025*(atan(abs(x[adj]))*180/pi)+1)*N)

proposed by J.R. Eastman, it measures abstract cost; slope in degrees. See: Vaissie E., Mobility of Paleolithic Populations: Biomechanical Considerations and Spatiotemporal Modelling, in PaleoAnthropology 2021 (1): 120-144 (with previous reference to Eastman 1999).

Pandolf et al.'s metabolic energy expenditure cost function (in Watts):

1 / ((1.5 * W + 2.0 * (W + L) * (L / W)^2 + N * (W + L) * (1.5 * (V^2) + 0.35 * V * (abs(x[adj])*100)))*N)

where W is the walker's body weight (Kg), L is the carried load (in Kg), V is the velocity in m/s, N is a coefficient representing ease of movement on the terrain (see above). Note that if V is set to 0 by the user, it is internally worked out on the basis of the Tobler function for on-path hiking; therefore, V will not be considered constant throughout the analysed area, but will vary as function of the slope.

For this cost function, see Pandolf, K. B., Givoni, B., & Goldman, R. F. (1977). Predicting energy expenditure with loads while standing or walking very slowly. Journal of Applied Physiology, 43(4), 577-581. https://doi.org/10.1152/jappl.1977.43.4.577.

For the use of this cost function in a case study, see Rademaker, K., Reid, D. A., & Bromley, G. R. M. (2012). Connecting the Dots: Least Cost Analysis, Paleogeography, and the Search for Paleoindian Sites in Southern Highland Peru. In White D.A. & Surface-Evans S.L. (Eds.), Least Cost Analysis of Social Landscapes. Archaeological Case Studies (pp. 32-45). University of Utah Press; see also Herzog, I. (2013). Least-cost Paths - Some Methodological Issues, Internet Archaeology 36 (http://intarch.ac.uk/journal/issue36/index.html) with references. For the idea of using an hiking function inside an energetic cost function, see for instance White D.A., Prehistoric Trail Networks of the Western Papaguarie. A Multifaceted Least Cost Graph Theory Analysis. In White D.A. & Surface-Evans S.L. (Eds.), Least Cost Analysis of Social Landscapes. Archaeological Case Studies (pp. 188-206). University of Utah Press.

Note: in the returned charts, the cost is transposed from Watts to Megawatts (see, e.g., Rademaker et al 2012 cited above).

Pandolf et al.'s metabolic energy expenditure cost function with correction factor for downhill movements (in Watts):

ifelse(abs(x[adj])*100 > 0 , 1 / (1.5 * W + 2.0 * (W+L) * (L/W)^2 + N * (W+L) * (1.5 * V^2 + 0.35 * V * (abs(x[adj])*100))), 1 / ((1.5 * W + 2.0 * (W+L) * (L/W)^2 + N * (W+L) * (1.5 * V^2 + 0.35 * V * (abs(x[adj])*100))) - (N * ((abs(x[adj])*100) * (W+L) * V/3.5) - ((W+L) * ((abs(x[adj])*100)+6)^2/W) + (25-V^2))))

for the parameters W, L, V, and N, see above. If V is set to 0 by the user, it is internally worked out on the basis of the Tobler function for on-path hiking; therefore, V will not be considered constant throughout the analysed area, but will vary as function of the slope. For the correction factor applied to the Pandolf et al.'s cost function, see Yokota M., Berglund L.G., Santee W.R., Buller M.J., Hoyt R.W. (2004), Predicting Individual Physiological Responsed During Marksmanship Field Training Using an Updates Scenario-J Model. U.S. Army Research Institute of Environmental Medicine Technical Report T04-09. For an archaeological application of the Pandol et al's cost function with correction factor, see White D.A., Prehistoric Trail Networks of the Western Papaguarie. A Multifaceted Least Cost Graph Theory Analysis. In White D.A. & Surface-Evans S.L. (Eds.), Least Cost Analysis of Social Landscapes. Archaeological Case Studies (pp. 188-206). University of Utah Press.

Note: in the returned charts, the cost is transposed from Watts to Megawatts (see, e.g., Rademaker et al 2012 cited above).

Minetti et al.'s metabolic energy cost function (in J/(kg*m)):

1 / (((280.5 * abs(x[adj])^5) - (58.7 * abs(x[adj])^4) - (76.8 * abs(x[adj])^3) + (51.9 * abs(x[adj])^2) + (19.6 * abs(x[adj])) + 2.5) * N)

see Minetti A.E., Moia C., Roi G.S., Susta D., Ferretti G. (2002), Enery cost of walking and running at extreme uphilland downhill slopes, in Journal of Applied Physiology 93, 1039-1046. Note that this equation is valid for slopes in the range -0.5/0.5; outside this range, the function's output becomes counterintuitive, as noted in Paez et al. in Journal of Transport Geography 82 (2020) and Herzog I. (2013), Theory and practice of cost functions, in Contreras F., Farjas M., Melero F.J. (eds), "Fusion of cultures. Proceedings of the 38th annual conference on computer applications and quantitative methods in archaeology". BAR IS, 2494, 375-382. Oxford: Archaeopress. In the latter work, Herzog proposes to replace Minetti et al.'s equation with its 6th degrees polynomial approximation (see the Herzog's metabolic cost function below).

Herzog's metabolic cost function in J/(kg*m):

1 / (((1337.8 * abs(x[adj])^6) + (278.19 * abs(x[adj])^5) - (517.39 * abs(x[adj])^4) - (78.199 * abs(x[adj])^3) + (93.419 * abs(x[adj])^2) + (19.825 * abs(x[adj])) + 1.64) * N)

see Herzog, I. (2016). Potential and Limits of Optimal Path Analysis. In A. Bevan & M. Lake (Eds.), Computational Approaches to Archaeological Spaces (pp. 179-211). New York: Routledge. Herzog suggests to use this as a 6th degree polynomial approximation of the Minetti et al's cost function (see above).

Van Leusen's metabolic energy expenditure cost function (in Watts):

1 / ((1.5 * W + 2.0 * (W + L) * (L / W)^2 + N * (W + L) * (1.5 * (V^2) + 0.35 * V * ((abs(x[adj])*100) + 10)))*N)

which modifies the Pandolf et al.'s equation; see Van Leusen, P. M. (2002). Pattern to process: methodological investigations into the formation and interpretation of spatial patterns in archaeological landscapes. University of Groningen. Note that, as per Herzog, I. (2013). Least-cost Paths - Some Methodological Issues, Internet Archaeology 36 (http://intarch.ac.uk/journal/issue36/index.html) and unlike Van Leusen (2002), in the above equation slope is expressed in percent and speed in m/s; also, in the last bit of the equantion, 10 replaces the value of 6 used by Van Leusen (as per Herzog 2013).
As explained above, if V is set to 0 by the user, it is internally worked out on the basis of the Tobler function for on-path hiking; therefore, V will not be considered constant throughout the analysed area, but will vary as function of the slope.

Note: in the returned charts, the cost is transposed from Watts to Megawatts.

Llobera-Sluckin's metabolic energy expenditure cost function (in KJ/m):

1 / ((2.635 + (17.37 * abs(x[adj])) + (42.37 * abs(x[adj])^2) - (21.43 * abs(x[adj])^3) + (14.93 * abs(x[adj])^4)) * N)

for which see Llobera M.-Sluckin T.J. (2007). Zigzagging: Theoretical insights on climbing strategies, Journal of Theoretical Biology 249, 206-217.

Ardigo et al.'s metabolic energy expenditure cost function (in J/(kg*m)):

1 / ((1.866 * exp(4.911*abs(x[adj])) * V^2 - 3.773 * exp(3.416*abs(x[adj])) * V + (45.71*abs(x[adj]^2) + 18.90*abs(x[adj])) + 4.456) * N)

see Ardigo L.P., Saibene F., Minetti A.E. (2003), The optimal locomotion on gradients: walking, running or cycling?, in Eur J Appl Physiol 90, 365-371. If V is set to 0 by the user, it is internally worked out on the basis of the Tobler function for on-path hiking; therefore, V will not be considered constant throughout the analysed area, but will vary as function of the slope.

Hare's metabolic energy expenditure cost function (in cal/km):

1 / ((48+30/(1/(6 * exp(-3.5 * abs(x[adj] + 0.05)))))*N)

see Hare T.S., Using Measures of Cost Distance in the Estimation of Polity Bounrdaries in the Post Classic Yautepec valley, Mexico, in Journal of Archaeological Science 31 (2004). Energetic expenditure is expressed in calories per km; walking speed is internally worked out from the DTM using the on-path Tobler's hiking function, which is expressed as its reciprocal; walking speed in km/h as per original Tobler's equation and as requested by Hare's function.

Note that the walking-speed-related cost functions listed above are used as they are, while the other functions are reciprocated. This is done since "gdistance works with conductivity rather than the more usual approach using costs"; therefore "we need inverse cost functions" (Nakoinz-Knitter (2016). "Modelling Human Behaviour in Landscapes". New York: Springer, p. 183). As a consequence, if we want to estimate time, we have to use the walking-speed functions as they are since the final accumulated values will correspond to the reciprocal of speed, i.e. pace. In the other cases, we have to use 1/cost-function to eventually get cost-function/1.

Value

The function returns a list storing the following components

• dtm: Digital Terrain Model ('RasterLayer' class)

• cost.surface: raster representing the cost-surface ('RasterLayer' class)

• accumulated.cost.raster: raster representing the accumualted cost ('RasterLayer' class)

• isolines: contour lines derived from the accumulated cost surface ('SpatialLinesDataFrame' class)

• LCPs: estimated least-cost paths ('SpatialLinesDataFrame' class)

• LCPs.back: estimated least-cost paths back to the origin ('SpatialLinesDataFrame' class)

• LCPs$length: length of each least-cost path (units depend on the unit used in the input DTM)

• LCPs.back$length: length of each least-cost path back to the origin (units depend on the unit used in the input DTM)

• dest.loc.w.cost: copy of the input destination location(s) dataset with a new variable ('cost') added; if the cost is expressed in terms of time, the 'cost' variable will store the time values in decimal numbers, while another variable named 'cost_hms' will store the time values in sexagesimmal numbers (hours, minutes, seconds)

• conductance: conductance 'Transitional Layer', returned because internally used by the movenetw() function

See Also

get_elev_raster, movecorr, movebound, movealloc, movecomp, movenetw, moverank

Examples

# load a sample Digital Terrain Model
data(volc)

# load a sample start location on the above DTM
data(volc.loc)

# load the sample destination locations on the above DTM
data(destin.loc)

# calculate walking-time isochrones based on the on-path Tobler's hiking function (default),
# setting the time unit to hours and the isochrones interval to 0.05 hour;
# also, since destination locations are provided,
# least-cost paths from the origin to the destination locations will be calculated
# and plotted; 8-directions move is used

result <- movecost(dtm=volc, origin=volc.loc, destin=destin.loc, move=8, breaks=0.05)


# same as above, but using the Irmischer-Clarke's hiking function (male, on-path)

result <- movecost(dtm=volc, origin=volc.loc, destin=destin.loc, funct="icmonp",
move=8, breaks=0.05)


# same as above, but using the 'cognitive slope'

result <- movecost(dtm=volc, origin=volc.loc, destin=destin.loc, funct="icmonp",
move=8, breaks=0.05, cogn.slp=TRUE)


# calculate accumulated cost surface and the least-cost path between the
# origin and one destination, and also calculate the LCP back to the origin

results <- movecost(dtm=volc, origin=volc.loc, destin=destin.loc[2,], move=8, return.base = TRUE)

movecost documentation built on April 12, 2023, 12:37 p.m.