@envisim/geosampling

A TypeScript library with functions for spatial sampling.

This package provides functions for various spatial sampling techniques.

Installation

npm install @envisim/geosampling

Usage

Package exposes the following entry points:

• @envisim/geosampling/sample-continuous
• @envisim/geosampling/sample-finite
• @envisim/geosampling/model-geometry
• @envisim/geosampling/collect-properties
• @envisim/geosampling/select-intersects
• @envisim/geosampling/point-processes
• @envisim/geosampling/errors

Sampling from continuous populations

You can sample fixed-size area plots within a collection of areas defined in WGS84 coordinates (e.g., forest stands, administrative regions). The model geometry for the plots is defined in planar units (meters), and the library handles placing them correctly onto the WGS84 map.

import { Feature, FeatureCollection, Polygon } from "@envisim/geojson";
import { rectangularAreaGeometry } from "@envisim/geosampling/model-geometry";
import { sampleAreaFeaturesOnAreas } from "@envisim/geosampling/sample-continuous";

// 1. Define the population area (sampling frame)
const polygon = Polygon.create([
  [
    [0, 0],
    [1, 0],
    [1, 1],
    [0, 1],
    [0, 0],
  ],
]);

const frame = FeatureCollection.newArea([new Feature(polygon, {})]);

// 2. Define the sampling unit's model geometry
const modelGeometry = rectangularAreaGeometry(10.0);

// 3. Select the sample
sampleAreaFeaturesOnAreas(frame, {
  pointSelection: "independent",
  sampleSize: 10,
  modelGeometry,
});

Explanation:

1. We create a FeatureCollection containing the area where sampling will occur, using WGS84 longitude and latitude coordinates.
2. We define the shape and size (model geometry) of the individual sample plots using functions like rectangularAreaGeometry. These dimensions are specified in meters.
3. The sampleAreaFeaturesOnAreas function takes the WGS84 population areas and the meter-based model geometry options. It calculates the placement of the sample plots onto the WGS84 map, returning a new FeatureCollection containing the sampled plots with their coordinates now in WGS84. Each feature includes properties like _designWeight which might be useful for subsequent statistical analysis.

Sampling from finite populations

import { Feature, FeatureCollection, Point } from "@envisim/geojson";
// Import functions to define model geometry (in meters) and perform sampling
import { sampleSpatiallyBalanced } from "@envisim/geosampling/sample-finite";

// 1. Define the sampling frame
const frame = FeatureCollection.newPoint([
  Point.create([0, 1]),
  Point.create([1, 2]),
  Point.create([2, 3]),
  Point.create([3, 4]),
  Point.create([4, 5]),
]);

// 2. Select the sample
sampleSpatiallyBalanced(frame, {
  method: "lpm2",
  sampleSize: 2,
  spreadOn: [],
  spreadGeo: true,
});

Model geometries

import { rectangularAreaGeometry } from "@envisim/geosampling/model-geometry";
// Create a 10 x 10 meter model geometry
rectangularAreaGeometry(10.0);

Collect properties from collections

import { collectProperties } from "@envisim/geosampling/collect-properties";

Create intersect of collections

import { selectAreaIntersectsArea } from "@envisim/geosampling/select-intersects";

Point processes

import { FeatureCollection, Polygon } from "@envisim/geojson";
import { uniformPoissonPointProcess } from "@envisim/geosampling/point-processes";

const frame = FeatureCollection.newArea([]);
frame.addGeometry(
  Polygon.create([
    [
      [0, 0],
      [1, 0],
      [1, 1],
      [0, 1],
      [0, 0],
    ],
  ]),
  {},
);

// Sample points uniformly on the frame, with an intensity of .00001 points per square meter.
uniformPoissonPointProcess(frame, { intensity: 0.00001 });

Error code lists

Error code lists of geosampling procedures.