LAMP is a system designed for local use to provide detailed short range weather guidance forecasts. The system provides hourly forecasts for up to 20 projections at all locations currently receiving Model Output Statistics (MOS) guidance, as well as for many additional sites within a National Weather Service Forecast Office's area of responsibility. LAMP can be initialized at any hour to provide guidance to support any forecast release time.

LAMP consists of a series of programs to analyze hourly data, run simple numerical models, evaluate statistical forecast equations, and display the information. The statistical forecasts are produced by regression equations that combine information from the central MOS guidance, local model output, and the most recent hourly surface observations. LAMP guidance updates MOS each hour by using the latest observational information and also furnishes forecasts for hourly projections for all stations in the coterminous United States with a record of hourly data.

In addition to providing forecast guidance, LAMP objectively analyzes the surface observations each hour. The gridfields produced by the analyses, and fields derived by them, can be displayed for the forecaster and are used to provide predictors for the statistical guidance. The local analyzes are also used to initialize the three, locally run models used in LAMP.

The LAMP analysis grid covers most of the coterminous United States and Canada. LAMP uses a successive correction analysis technique on quasi-continuous variables. Temperature, dew point, wind (U and V components and speed), sea level pressure, and a moisture variable known as the saturation deficit are analyzed by this method. The saturation deficit indicates the degree of saturation in the 1000- to 500-mb layer. A nearest neighbor analysis scheme is used for non-continuous variables, such as cloud amount.

The three simple numerical models used in LAMP provide advective forecasts of sea level pressure, saturation deficit, and variables such as surface temperature and cloud amount. These models are driven by 500-mb height forecasts obtained from the National Meteorological Center's Nested Grid Model (NGM). These 500-mb heights are interpolated, both in space and time, to obtain an estimate at each hour over the LAMP grid.

Two equation sets are derived for each start time-warm season equations valid for the months of April through September, and cool season equations valid for the remainder of the year. These equations are based on approximately 10 seasons of data.

AWIPS will provide the capability for models to be run locally to provide guidance to forecasters over a forecast office's area of responsibility. LAMP is such a model and will produce forecast guidance in hourly steps out to 20 hours from any initialization time. These forecasts can be used to help initialize the Interactive Computer Worded Forecast, from which many products can be automatically generated such as public zone and agricultural forecasts. Support to aviation will include forecasts of ceiling height, visibility, and the amount and height of up to three layers of clouds.

This forecast guidance can currently be displayed in the form of an areal map displaying various weather forecast elements. In addition, it is possible to display this areal information contoured or as a standard station plot. Here is an excellent example of a LAMP display in AWIPS showing various meteorological images including saturation deficit, temperature/SLP, and station plots.

By individual station, the user has the option of displaying a time series, a comparison of LAMP vs. NGM MOS (in bulletin format), and a threshold diagram. These indicate probabilities of and thresholds for low visibility and precipitation occurring on the hour (a best category forecast of low visibility and/or precipitation occurring is made when the probability exceeds the threshold).

For a more detailed description of the LAMP model, click HERE.