The 2nd day of the IESWG meeting covered several Global and Regional Land Data Assimilation Systems their overviews and recent progress.

Few key outcomes:

Unification of LDAS methodology and system has greater sustainability, and is a target highlighted both in UKMO and ECMWF presentations.

L1 to L4 there is a variety of applications, starting from L4 allow easier access to LDAS work, L1 is appealing for Coupled DA?

What can make LST products be more readily useable? Can it be used in LDAS? Are models good in enough?

NRT access to future Copernicus Expansion Mission and access to BUFR files for L1 data.

In general to satellite data providers NRT processing capability of L2 to L4 increase the uptake by operational Centres. 

There is some inertia in exploring new products eg. for snow IMS snow cover is well established, what about SWE or other SC products (eg. H-SAF)? A solution is to involve early the LDAS teams?

Land Cover and Land Cover change and soil texture are critical ancillary that all modellers need get right and may facilitate LDAS activities. Ultimately reconsidering the role of ancillary.

Urban dataset 

2DVAR 2DEnVAR show adtantages over the OI and can more easily accomodate satellite data (starting from ASCAT) as shown in DWD presentation.

New SnoTEL network insitu snow depth observation entering GTS, DWD showed benefits (1000 more observations) and challenges for the DWD LDAS. COOP observations should be in the pipeline for being shared.

JMA showed encouraging results from satellite data assimilation of AMSU data.

Assimilating VOD can improve NWP (shown by Pete Weston at ECMWF).

This may call to greater attention to Dynamid Vegetation modelling (also in NWP context)?

Irrigation might be a connected modelling need?

NCMRWF LDAS make use of ASCAT/screenlevel and India insitu network (Agro-AWS) for soil moisture and soil temperature very useful for verification. Model variability inferior to observations variability.

NOAA-NESDIS working on a set of ET products over CONUS and Global domains.

CCI-LST working on a 20+ year 3-hourly dataset for IR intercalibrated Land Surface Temperature from GEO and LEO satellites.