diff --git a/packages/k-means/src/kmeans.ts b/packages/k-means/src/kmeans.ts
index b9778306f3..f1bf9236ee 100644
--- a/packages/k-means/src/kmeans.ts
+++ b/packages/k-means/src/kmeans.ts
@@ -4,6 +4,18 @@ import { SYSTEM, uniqueIndices, weightedRandom } from "@thi.ng/random";
 import { add, median, mulN, ReadonlyVec, Vec, zeroes } from "@thi.ng/vectors";
 import type { CentroidStrategy, Cluster, KMeansOpts } from "./api";
 
+/**
+ * Takes an array of n-dimensional `samples` and attempts to assign them to `k`
+ * clusters, using the behavior defined by (optionally) given `opts`.
+ *
+ * @remarks
+ * https://en.wikipedia.org/wiki/K-medians_clustering
+ *
+ * @param k
+ * @param samples
+ * @param opts
+ * @returns
+ */
 export const kmeans = <T extends ReadonlyVec>(
     k: number,
     samples: T[],
@@ -45,7 +57,8 @@ export const kmeans = <T extends ReadonlyVec>(
 };
 
 /**
- * k-means++ initialization / selection of initial cluster centroids.
+ * k-means++ initialization / selection of initial cluster centroids. Default
+ * centroid initialization method for {@link kmeans}.
  *
  * @remarks
  * References:
@@ -113,6 +126,12 @@ const buildClusters = (centroids: ReadonlyVec[], clusters: number[]) => {
     return indices;
 };
 
+/**
+ * Default centroid strategy forming new centroids by averaging the position of
+ * participating samples.
+ *
+ * @param dim
+ */
 export const means: CentroidStrategy = (dim) => {
     const acc = zeroes(dim);
     let n = 0;
@@ -125,6 +144,12 @@ export const means: CentroidStrategy = (dim) => {
     };
 };
 
+/**
+ * Centroid strategy forming new centroids via componentwise medians.
+ *
+ * @remarks
+ * https://en.wikipedia.org/wiki/K-medians_clustering
+ */
 export const medians: CentroidStrategy = () => {
     const acc: ReadonlyVec[] = [];
     return {
@@ -132,3 +157,41 @@ export const medians: CentroidStrategy = () => {
         finish: () => (acc.length ? median([], acc) : undefined),
     };
 };
+
+/**
+ * Means centroid strategy for decimal degree lat/lon positions (e.g. WGS84).
+ * Unlike the default {@link means} strategy, this one treats latitude values
+ * correctly in terms of the ±180 deg boundary and ensures samples on either
+ * side of the Pacific are forming correct centroids.
+ *
+ * @remarks
+ * When using this strategy, you should also use the
+ * {@link @thi.ng/distance#HAVERSINE_LATLON} distance metric for
+ * {@link KMeansOpts.distance}.
+ *
+ * @example
+ * ```ts
+ * kmeans(3, [...], { strategy: meansLatLon, dist: HAVERSINE_LATLON })
+ * ```
+ *
+ * https://en.wikipedia.org/wiki/World_Geodetic_System
+ */
+export const meansLatLon: CentroidStrategy = () => {
+    let lat = 0;
+    let lon = 0;
+    let n = 0;
+    return {
+        update: ([$lat, $lon]) => {
+            lat += $lat < 0 ? $lat + 360 : $lat;
+            lon += $lon;
+            n++;
+        },
+        finish: () => {
+            if (!n) return;
+            lat /= n;
+            if (lat > 180) lat -= 360;
+            lon /= n;
+            return [lat, lon];
+        },
+    };
+};