solar_PV {GenEst}

Photovoltaic Example Dataset

Description

An example data set for estimating fatalities from a large photovoltaic solar generation facility.

The simulated site is organized into 300 arrays of panels. As observers walk north-south along paths between arrays, they look east or west down rows between solar panels 150 meters long, with 38 searchable rows per array. Observers consistently look for animals down one cardinal direction, making this a one-sided distance sample. Searches are scheduled on a seven day rotation, with 60 arrays searched per weekday. A sitewide clearout search is implemented before the first scheduled winter search.

The dataset consists of five parts: Data on carcass observations (CO) from array searches, field trials for estimating carcass persistence (CP) and searcher efficiency (SE), search schedule (SS), and density weighted proportion (DWP) of area searched at each array (which is an area adjustment factor to account for incomplete search coverage).

Usage

solar_PV

Format

solar_PV is a list with 5 elements:

SE

Searcher efficiency trial data

CP

Carcass persistence trial data

SS

Search schedule parameters

DWP

Density weighted proportion of area searched

CO

Carcass observations

Searcher Efficiency (SE)

$SE is a data frame with each row representing the fate of a single carcass in the searcher efficiency trials. There are columns for:

Season

"winter", "spring", "summer", or "fall"

Size

"bat"; or "lrg", "med", or "sml" bird

"Search1",...,"Search5"

fate of carcass on the 1st, 2nd, 3rd, 4th, and 5th search after placement. A value of 1 implies that a carcass was discovered by searchers, 0 implies the carcass was present but not discovered, and any other value is interpreted as "no search" or "carcass not present" and ignored in the model. In this data set, NA indicates that a carcass had been previously discovered and removed from the field. A user may use a variety of values to differentiate different reasons no search was conducted or the carcass was not present. For example, "NS" to indicate the search was not scheduled in that location at that time, or "SC" to indicate the carcass had been removed by scavengers prior to the search.

Distance

the distance a carcass was placed from the observer's transect. Used in determining probability to detect with distance sampling.

Carcass Persistence (CP)

$CP is a data frame with each row representing the fate of a single carcass in the carcass persistence trials. There are columns for:

Index

unique ID for each carcass

Season

"winter", "spring", "summer", or "fall"

Size

"bat"; or "lrg", "med", or "sml" bird

LastPresent, FirstAbsent

endpoints of the interval bracketing the time the carcass was scavenged or otherwise removed from the field. For example, LastPresent = 2.04, FirstAbsent = 3.21 indicates that the carcass was last observed 2.04 days after being placed in the field and was noted missing 3.21 days after being placed. If the precise time of carcass removal is known (e.g., recorded by camera), then LastPresent and FirstAbsent should be set equal to each other. If a carcass persists beyond the last day of the field trial, LastPresent is the last time it was observed and FirstAbsent is entered as Inf or NA.

Search Schedule (SS)

$SS is a data frame with a row for each date an array at the site was searched, a column of SearchDates, and a column for each array. In addition, there is an optional column to indicate the Season. The columns for distinct area (array) and the date column are required, and the names of the columns for search areas must match the names of areas used in the DWP and CO files.

SearchDate

columns of dates when arrays were searched. Format in this data is "%Y-%m-%d CDT", but time zone (CDT) is optional. A time stamp may be included if desired (e.g., 2018-03-20 02:15:41). Alternatively, \ can be used in place of -.

Season

"winter", "spring", "summer", or "fall" to indicate which season the search was conducted in. Season is optional but may be used as a temporal covariate for fatality estimates.

Density Weighted Proportion (DWP)

$DWP is a data frame with a row for each array and columns for each carcass size class (labels must match those of the class factors in the carcass observation file). Values represent the density-weighted proportion of the searched area for each size (or the fraction of carcasses that fall in the searched area). In this example, observers walk along transects separated by 150 meters, and search coverage is assumed to be 100 DWP = 1 for each unit. This requires that carcasses be placed at random locations in the field, even at distances from the transects that would make it unlikely to observe small carcasses.

Unit

unique ID for each array. IDs match those used in the $CO data frame and the column names in the $SS data.

bat

DWP associated with size class Bat

sml

DWP associated with size class Small

med

DWP associated with size class Medium

lrg

DWP associated with size class Large

Carcass Observations (CO)

$CO is a data frame with a row for carcass observed in the carcass searches and a number of columns giving information about the given carcass (date found, size, species, etc.)

Index

unique identifier for each carcass.

Unit

identifier for which unit the given carcass was found at: "arc19", "arc65", etc, for arcs in the outer heliostat field, or "center", indicating the inner heliostat field.

Species

species of the carcass: "BA", "BB", "BC", "BD", "BE", "LA", "LB", "LD", "LE", "MA", "MB", "SA", "SB", "SC", "SD", "SE", "SF", "SG"

Size

size: "bat", "lrg", "med", "sml"

Row

Optional indicator of which row within an array a carcass was found at.

Distance

The perpendicular distance from the searcher's transect at which the carcass was discovered at.

DateFound

dates entered in the same format as in $SS$SearchDate. Every date entered here is (and must be) included in the search schedule ($SS$SearchDate

X

UTM Easting of carcass.

Y

UTM Northing of carcass.

Source

solar_PV