# README file for (sub)Arctic DMS emission dataset (latitudes >45N), v1.0.0
===========================================================================

Martí Galí Tàpias, 2019-06-12
http://orcid.org/0000-0002-5587-1271

Address requests for additional data and questions to:
marti.gali.tapias@gmail.com

This dataset contains satellite-derived time series of:
* dmspt (sea-surface dimethylsulfoniopropionate -DMSPt- concentration, in µmol/m3 or 
* dms (sea-surface dimethylsulfide -DMS- concentration, in µmol/m3 or equivalent nmol/L)
equivalent nmol/L)
* fdms (sea-air DMS flux, in µmol/m2/d)
* ice concentration (0 to 1)
* additional dataset with curated in situ data used for validation

The satellite DMSPt dataset was produced as described by:
Galí, M., Devred, E., Levasseur, M., Royer, S. J., & Babin, M. (2015). A remote sensing algorithm for planktonic dimethylsulfoniopropionate (DMSP) and an analysis of global patterns. Remote Sensing of Environment, 171, 171-184. https://doi.org/10.1016/j.rse.2015.10.012

The satellite DMS dataset was produced as described by:
Galí, M., Levasseur, M., Devred, E., Simó, R., & Babin, M. (2018). Sea-surface dimethylsulfide (DMS) concentration from satellite data at global and regional scales. Biogeosciences, 15(11), 3497-3519. https://doi.org/10.5194/bg-15-3497-2018

Galí, M., Devred, E., Babin, M., & Levasseur, M. (2019). Decadal increase in Arctic dimethylsulfideemission. In revision

Code for the DMSPt_SAT and DMS_SAT algorithms: https://github.com/mgali/DMS-SAT_ALGORITHM


## File format

NetCDF files are self-describing: they contain one “variable of interest” and all the corresponding temporal and geographic dimensions.

## File naming convention

variable_sensorYYYY_nD_28km.nc

Both the DMSPt and DMS satellite datasets are based on ocean color data obtained from NASA Ocean Color website. Chlorophyll was derived from the GSM algorithm, hence _gsm was appended to the variable names.

DMSPt and DMS were produced for 2 satellite sensors.
* S = SeaWiFS, years 1998 to 2007
* A = MODIS-Aqua, years 2003 to 2016 (the last year is empty from October on)

FDMS was calculated using ERA-Interim wind speed and SST, and ice cover from NSIDC. Two different gas exchange parameterizations were used, whose names are appended to the variable names:
* W97 (Woolf 1997)
* N00GBC (Nightingale et al. 2000 GBC –do not confound with Nightingale et al. 2000 GL)

FDMS was also calculated using climatological DMS fields (W97 parameterization) while retaining the interannual variability of all the other variables involved in FDMS calculation:
* S19982007 DMS climatology
* A20032016 DMS climatology
* L11 DMS climatology (interpolation based climatology of Lana et al. 2011 GBC)

Ice cover (ice) is also provided for the full period 1998-2016.

## Directory naming convention

Each directory (zipped) is named using the names of variable it contains


## Additional dataset with in situ data used for validation

This dataset (provided in both *.csv and *.mat formats) contains quality-controlled in situ measurements mostly obtained from the NOAA PMEL database, plus additional data provided by collaborators**. Data are binned to 8-day 28 km resolution matching the satellite datasets, and span the same period (1998-2016). Coordinates correspond to pixel coordinates. We report means, range, sd and counts for DMS, DMSPt and Chl within each documented bin. For general quality control procedures see:

https://github.com/mgali/DMS-SAT_DATA_DEV_VAL

**We thank Martine Lizotte, Guillaume Massé, Sohiko Kameyama, Tereza Jarniková and Philippe Tortell