Validation of AMSR-E Polar Ocean Products Using a Combination of Modeling and Field Observations

Investigators:

Jim Maslanik, Julienne Stroeve, Thorsten Markus, Matthew Sturm and John Heinrichs

Contact: james.maslanik@colorado.edu

303-492-8974

Goals

Complement the AMSR Science Team’s validation plan by:

collecting in-situ "vertical column" evaluation data

obtaining basic in-situ data over extended areas

extending validation to poorly sampled conditions

quantifying error sources via R/T modeling

Objectives

Quantify the error characteristics of sea ice-related products as a function of a range of atmospheric and surface conditions through direct comparison of data and through simulations;

Generate a suite of model-derived, simulated AMSR radiances for a range of observed and theoretical combinations of surface and atmospheric conditions for use in validation assessment;

Consider factors related to consistency of the long-term time series of sea ice data;

Document improvements achieved by AMSR through comparisons of AMSR- and SSM/I-derived products;

Summarize product accuracies in ways that are directly relevant to users seeking to understand how well the products are likely to perform for their particular applications.

This work extends the Science Team validation plan by:

Providing necessary in situ measurements of snow and ice conditions, providing high-resolution surface mapping below cloud cover, and providing atmospheric profile observations co-located with P-3 overflights and also during periods not covered by the Science Team aircraft missions;

Use of radiative transfer modeling to determine the physical conditions that are most likely to account for observed product errors and differences between the two sea ice algorithms;

Rationale

Data and modeling are needed to provide an improved physical basis for interpreting differences between the two AMSR ice algorithms;

Methods to reduce weather contamination and coastal effects differ between the two algorithms, and require specific evaluation;

Snow depth and ice temperature products are new and require comprehensive validation data over extended areas (i.e., long transects and low-level aircraft mapping patterns);

Evaluation data needed for cloudy as well as clear conditions, and during conditions that preclude surface observations using piloted aircraft;

New applications require more extensive understanding of error magnitudes and causes.

Approach

(1) Collect in-situ surface and atmospheric data needed to directly validate products and to diagnose and assess algorithms improvements and sensitivities; and

(2) Use these data to calculate algorithm sensitivities to the range of conditions introduced by variations in surface and atmospheric conditions.

Tools

1. Field Instrumentation

CRREL "transect" instruments (snow/ice interface temperatures, snow depth probe, sled-mounted radar)

Sled-mounted radiometric instruments

Aerosonde Unpiloted Aerial Vehicles (UAV)

2. Modeling

MWMOD

MEMLS (Microwave Emission Model of Layered Snowpacks)

SNTHERM

Components of the Data Collection Plan

In situ measurements and UAV observations of fast ice in the Barrow area;

In situ measurements in Elson Lagoon and freshwater lakes in the Barrow area;

UAV observations of pack ice conditions in the Beaufort and Chukchi seas;

UAV observations of atmospheric conditions over the fast ice and pack ice areas;

Assembly of atmospheric measurements obtained from the Barrow DOE ARM site, the NOAA CMDL site, and satellite atmospheric sounders.

Key elements of the methodology

Coordination with already-planned and funded field data collection efforts to provide information most useful to the validation effort. The proposed work takes advantage of these previously-funded projects to provide a wealth of data at relatively little additional cost to NASA;

Extension of these planned field efforts through additional data collection targeted specifically at AMSR validation needs, including additional in situ measurements and use of robot aircraft to provide unique, detailed mapping of surface conditions and atmospheric profiles.

Use of radiative modeling to assess algorithm accuracies given different surface and atmospheric conditions (sensitivity studies, error bounds);

Forward radiance modeling as an additional validation tool, with the primary benefit of synthesizing multiple types of validation data into simulated top-of-the-atmosphere radiances that can be directly compared to AMSR and P-3 radiances;

Comparison of the AMSR products to SSM/I data to assess the degree of improvement provided by the new products and to quantify issues of consistency for extension of the SMMR-SSM/I time series.

Data Sets

in situ measurements of snow properties, snow depth, temperatures, ice thickness, ice type, and ice characteristics along transects in the fast ice zone in the Barrow area and over a large saltwater embayment (Elson Lagoon) near Barrow;

UAV-acquired surface mapping (ice concentration and type), skin temperatures, surface roughness, and vertical profiles of atmospheric temperature, pressure, and humidity from approximately 100 m to 6000 m altitude. UAV observations will be coincident with in situ transects, measurements from the Healy, flights of the NASA P-3, and EOS Aqua overpasses covering portions of the Beaufort and Chukchi seas and the North Slope coast;

Assembly of atmospheric information obtained by the Barrow DOE ARM site and NOAA CMDL facilities, as well as additional profile information generated by satellite atmospheric sounders (TOVS and AIRS);

Synthetic AMSR-frequency data sets .

Advantages of the Aerosonde UAV

Long flight duration compared to piloted aircraft;

Can operate at low altitude and below the persistent Arctic stratus cloud layer;

Robotic nature, slow flight speed, and maneuverability allows for detailed and precise flight patterns;

Low risk and relatively low cost.