The Great Migration is not a date. It is a system. About 1.5 million wildebeest, 200,000 zebra, and 350,000 Thompson’s gazelle move in a long, slow, weather-driven circle through Tanzania and Kenya. The system has no leader and no schedule. But it does have inputs.
The two big inputs are rainfall and grass quality. The herds follow the rain — but with a one-to-three-week lag, because they will not move into wet grass until the wet grass is past its first sugar-poor phase. So we do not actually watch the rain. We watch the grass.
Our placement model uses NDVI satellite data (a measure of vegetation greenness, updated every five days) layered onto seven-day rainfall forecasts from the World Meteorological Organisation. We then overlay where the herds are sitting at any given moment, using ranger-network reports from inside the Serengeti and from across the Mara River in Kenya.
From that, we get a forecast of where the leading edge of the herds will be in roughly fourteen days. That is the window in which we move a mobile camp. Fourteen days is the minimum lead time we need to break down, transport, and re-pitch a fully-staffed camp.
It is not always right. The herds have, on more than one occasion, ignored the model entirely. But when it works — and it works often — you will sit on a folding chair, with a cold drink, and watch 12,000 wildebeest cross a river that no one knew they would cross until a Wednesday morning eleven days earlier.
Filed under
Field notes
