We are often asking a precipitation gauge that is measuring a 0.2 to 0.32 square meter sample to represent the precipitation of a 1000 square meter or larger geographic area. The spatial distribution of precipitation is spatially non-uniform and often random at the synoptic scale (over 1000 km). It can also be systematically non-uniform at the Meso-alpha scale (200-2000 km), the Meso-beta scale (20-200 km), the Meso-gamma scale (2-20 km), and at the Microscale (1-2 km or less). The real challenge is how to site precipitation gauges in order to obtain representative samples of precipitation events over such wide geographic areas.
This post is an excerpt from the Belfort Instrument Engineering Guide to Siting Precipitation Gauges. Click the button below to download the free guide now (PDF).
Precipitation is Not Uniformly Distributed
A good example of the non-uniform spatial distribution of precipitation is demonstrated by the radar depiction of a thunderstorm over relatively flat terrain in Oklahoma.
When measuring a randomly distributed phenomenon, a large number of point samples and statistical analysis are required to create an accurate area characterization. A single precipitation measurement slightly north or south of this cell would indicate little or no precipitation in the area. It would take a large network of hundreds of precipitation gauges, separated by as little as 1 or 2 km, to accurately characterize this precipitation event without the aid of Doppler Radar. Economic factors often dictate that the least number of gauges be used to characterize the precipitation over the desired area. Properly locating (siting) these gauges become a critical factor in characterizing the precipitation over such wide areas.
Terrain and Obstacles Can Introduce Systematic Errors to Precipitation Measurements
If a large number of gauges were sited without taking into account the systematic errors introduced by the terrain and other obstacles, a completely erroneous conclusion as to the total area precipitation could result.
For example, on a Mesoscale the impact of having more gauges sited on the windward side of a mountain than on the lee side of a mountain could lead to a conclusion that higher than actual total precipitation has occurred over an area inclusive of both sides of the mountain. Having more gauges sited in the lee of a large body of water could give a higher than actual estimate of total area precipitation. Having more gauges sited in the lee of a city complex could give a higher than actual estimate of total area precipitation. Having gauges located on the windward side of river valleys could give a higher than actual indication of total area precipitation. Understanding how Mesoscale obstructions and topography impact precipitation can lead users to site gauges at locations that don’t erroneously enhance or diminish precipitation estimates over wide geographic areas.
Microscale siting criteria deals primarily with wind as a local source of disturbance of the precipitation measurement. Obstacles of a uniform height that act as an effective wind break, but do not block precipitation, present the best sites for accurate precipitation measurement. One study showed measurably greater precipitation in forest clearings than on adjacent open farmland. Was this because the surrounding trees impacted wind speed and more precipitation fell into the forest gauge than in the open farmland or some other local phenomena? Siting gauges too close to trees, buildings and even small adjacent topographic anomalies may block blowing precipitation and significantly impact the gauges ability to accurately measure representative samples of total area precipitation. Siting gauges in open land without wind shielding can lead to lower than actual estimates of total area precipitation as the wind will blow the precipitation (especially solid precipitation at speeds in excess of 5 m/s) over the top of the gauge’s collection orifice. Siting gauges on top of buildings or tall poles where significantly higher wind speeds may occur may also lead to under catchment. Siting gauges in pits, level with the ground, can minimize wind induced errors but introduce over catchment due to measurement of drifting snow or debris. Precise guidelines for Microscale siting of precipitation gauges is essential for representative measurements.
This application note will focus on how these systematic errors introduced by siting choices at the Mesoscale and Microscale level can be minimized to provide representative precipitation gauge data for input into wider area precipitation estimates and analytical models.