Microwave Integrated Retrieval System (MIRS) - Geophysical Monitoring

The geophysical performance is the comparison of the MIRS products with other data sources.

Comparison to NWP ( GDAS and ECMWF )

GDAS:

This is done by using the Global Data Assimilation System (GDAS) analyses outputs. These gridded fields are first interpolated to the exact location of the satellite measurements before a point-to-point comparison is undertaken. It is also important to keep in mind that MIRS uses five (5) analyzes (00, 06, 12, 18, 24) to perform the time-interpolation of GDAS fields to avoid doing any extrapolation or using the 18h00 analysis for all data between 18h00 and 24h00. This requires that the geophysical performance at any given day is processed only two days after that particular day.

ECMWF:

This is to compare MIRS products with European Centre for Medium-Range Weather Forecasts (ECMWF) model outputs. For more information of ECMWF, please check ECMWF model website: ECMWF

Comparison to MSPPS

This is to compare MIRS advanced algorithm with heritage algorithm MSPPS. For more information of MSPPS, please check MSPPS website. .

Comparison with AMSR-E Ice and Snow Products

In order to assess the performance of the MIRS daily ice and snow retrieval products comparisons are made with standard products produced from Advanced Microwave Scanning Radiometer - Earth Observing System (AMSR-E) brightness temperatures using algorithms developed at NASA. The MIRS products that are compared are the total sea ice concentration (SIC) and snow water equivalent (SWE). In the case of sea ice, two independent AMSR-E products are used for assessment: the NASA Team 2, and the Bootstrap algorithm. For each MIRS product, several types of intercomparisons are presented. First, horizontal maps of each day's retrievals, differences between MIRS and the corresponding AMSR-E products, as well as scatterplots are shown. In the case of sea ice concentration, difference maps and scatterplots between the NASA Team 2 and Bootstrap are also provided to give an estimate of the uncertainty in the AMSR-E products themselves. The user is provided the option to display results on polar stereographic or global latitude/longitude grids. Second, time series containing daily values of various difference metrics and skill scores are presented. This alows for the assessment of performances over time. The user may display results for a single sensor product, or for all sensors simultaneously. Third, daily histograms comparing the MIRS with AMSR-E products are shown. For both the time series and histograms, results for northern and southern hemisphere sea ice concentration are displayed separately.

Comparison with SNODAS Snow Products

Daily comparisons are made of MIRS retrieved snow water equivalent (SWE) with the operational analyses produced from the NOAA National Weather Service National Operational Hydrologic Remote Sensing Center (NOHRSC) SNOw Data Assimilation System (SNODAS). The SNODAS analyses are produced once per day for the continental U.S. valid at 0600 UTC, and are obtained directly from NOHRSC. Assessments shown are maps, scatterplots, histograms and time series of performance metrics (e.g. bias, correlation, and Heidke skill score, etc.)

Comparison with SSMIS Sea Ice Products

Daily comparisons are made of MIRS retrieved total sea ice concentration (SIC) with daily near real-time retrievals of SIC from DMSP F17 SSMIS brightness temperatures using the NASA Team algorithm. The SSMIS near real-time estimates are produced on a 25 km polar stereographic grid for both northern and southern hemispheres and are obtained daily from the National Snow and Ice Data Center in Boulder, Colorado. Assessments shown are maps, scatterplots, histograms and time series of performance metrics (e.g. bias, correlation, and Heidke skill score, etc.)

Performance Time Series

The daily updated time series of temperature and water vapor mean bias and standard deviation is to help monitor MIRS performance.

Performance Tables

The daily updated table format of temperature and water vapor mean bias and standard deviation .