California merged satellite-derived 1-km dataset

Mati Kahru, mkahru@ucsd.edu    Updated : 8/16/2020

A merged 1-km dataset using data from multiple satellite sensors has been created for the California EEZ region. The methods are a combination of the methods described for the two related California Current datasets: the extended area 1-km dataset http://spg-satdata.ucsd.edu/ and the optimally merged 4-km dataset http://spg-satdata.ucsd.edu/CC4km/.

The datasets covers the state of California EEZ (Fig. 1A) on a grid of 1188 (width) x 1333 (height) with approximately 1 km step are in HDF4 format. The upper-left corner (lat, lon) is 42N; -129E; the lower-right corner is 30N; -115.791E. The longitude of each pixel can be calculated as Lon = -129 + 0.011119 * x where x is the column number from left to right (0 to 1187) and the latitude of each pixel can be calculated as Lat = 42 - 0.009002 * y where y is the row number from top to bottom (0 to 1332)

cid:image001.png@01D622D0.BAC2F850

Fig. 1. The EEZ boundaries (map in Albers Conic Equal Area projection) (left) and the ~1-km grid encompassing it (right).

A set of products are available for each variable. Various compositing operations are performed to fill the gaps caused by missing data (e.g. due to clouds). The following products are created (with slight modifications for individual variables):

(1) Daily merged dataset (Fig. 2, left for Chla) has typically many pixels with no data.

(2) Daily 5-day running mean. These are also daily estimates, centered at the middle day. The amount of missing data is reduced significantly(Fig. 2, right). The running mean operation is justified as Chla usually does not change drastically from day to day.

(3) 5-day composites (made from the daily running means).

(4) 5-day composites interpolated in time to further fill missing values. Note: only missing pixel values are affected.

(5) 5-day composites interpolated twice.  As Chla usually does not change rapidly over daily time scales (e.g. compared to Photosynthetically Available Radiation, PAR) these are used in daily Net Primary Production (NPP) estimates.

(6) Monthly composites in HDF4.

(7) Monthly composites in PNG.

 

Fig. 2. An example daily Chla image of 7-Jan-2020 (C2020007_chl_comp.remap_ca1km.hdf, left) and the corresponding 5-day running mean Chla (C20200052020009_chl_comp.hdf, right).

For Chla each pixel is unsigned byte in log10-scaling and the Chla concentration can be calculated from the pixel value (PV) as: Chl (mg m-3) = 10^(0.015 * PV - 2.0), i.e. 10 to the power of 0.015 * PV - 2.0. Pixel values 0 (black in Fig. 2) and 255 (white in Fig. 2) are considered invalid and must be excluded from any statistics. Also, PV = 1 may be used for annotating and for coastlines and has to be excluded too. The annotation color bar may be written over land into some of the products. When reading with Matlab the unsigned byte variable is sometimes reported as signed byte (int8, values from -127 to 128) and not as unsigned byte (values from 0 to 255) and values over 128 become negative. A simple fix is to add 256 if the signed pixel value is negative.

Surface chlorophyll-a concentration (Chla, mg m-3) is derived by mapping available Level-2 data of standard algorithms from multiple satellite sensors and merging the valid pixel values by averaging. Extensive validation has been done separately for each sensor and for the merged products (Fig. 3, e.g. Kahru et al., 2012). The merged Chla dataset starts with OCTS in November, 1996, and the continuously available data starts with SeaWiFS in September, 1997. Starting from 2000, data from more than one sensor have been merged. Using multiple sensors makes the datasets more reliable by increasing coverage and reducing the amount of missing data.

 

Fig. 3. Validation of the standard Chla algorithms for the sensors used in the merged Chla products. Match-ups with in situ Chla for MODIS-Aqua, MODIS-Terra, OLCI-A, OLCI-B, VIIRS-SNPP (VIIRSN) and VIIRS-NOAA20 (VIIRSJ) using in situ data from CalCOFI and CCE-LTER cruises (e.g. Kahru et al., 2012, 2015, 2018).

 

A standard optical parameter measuring water turbidity or the inverse of transparency is the attenuation of downwelling light at 490 nm, Kd490 or simply Kd (m-1).  Currently the Kd data are included not at 1 km resolution but at 4 km resoltion. Otherwise the grid is the same. The current Kd data are based on the ESA Ocean Colour Climate Change Initiative (OC-CCI) version 4.2 (Sathyendranath et al., 2019, https://esa-oceancolour-cci.org/).  The Lee et al. (2005) equation and bbw from Zhang et al. (2009) are used.

Net Primary Production (NPP, mg C m-2 day-1) estimates are calculated from the 5-day merged and interpolated (2x) Chla, merged daily PAR (currently from MODISA, MODIST, VIIRS-SNNP, VIIRS-JPSS1) and optimally interpolated daily sea-surface temperature (SST) data (Reynolds et al. 2007) using a modified VGPM (Behrenfeld and Falkowski 1997) called VGPM-CAL (Kahru et al. 2009). In contrast to ocean color variables, PAR retrieval is not limited by clouds and we have full daily PAR coverage during the satellite overpasses. Using 5-day Chla we are assuming that Chla is changing relatively slowly compared to PAR. By using daily PAR and daily SST (with 5-day Chla) we can capture the daily variability of NPP products even though we are using 5-day Chla (Kahru et al. 2016). NPP datasets are in HDF4 files where each pixel is an Int16 value of NPP in units of mg C m-2 day-1. Values <2 and >10,000 are invalid.

The datasets are in Zipped files and can be downloaded using the following hyperlinks:

Chla Daily

Chla Daily 5-day running mean

 

Chla 5-day and Monthly composites

1996, 1997, 1998

1996, 1997, 1998

 

Chla 5-day all years

1999, 2000, 2001

1999, 2000, 2001

 

Chla 5-day interpolated

2002, 2003, 2004

2002, 2003, 2004

 

Chla 5-day interpolated 2x

2005, 2006, 2007

2005, 2006, 2007

 

Chla monthly ;            Chla monthly PNG files

2008, 2009, 2010

2008, 2009, 2010

 

 

2011, 2012, 2013

2011, 2012, 2013

 

 

2014, 2015, 2016

2014, 2015, 2016

 

 

2017, 2018, 2019

2017, 2018, 2019

 

 

2020

2020

 

 

 

Kd Daily

Kd Daily 5-day running mean

 

Kd 5-day and Monthly composites

1996, 1997, 1998

1996, 1997, 1998

 

kd 5-day all years

1999, 2000, 2001

1999, 2000, 2001

 

kd 5-day interpolated

2002, 2003, 2004

2002, 2003, 2004

 

kd 5-day interpolated 2x

2005, 2006, 2007

2005, 2006, 2007

 

kd monthly ;            kd monthly PNG files

2008, 2009, 2010

2008, 2009, 2010

 

 

2011, 2012, 2013

2011, 2012, 2013

 

 

2014, 2015, 2016

2014, 2015, 2016

 

 

2017, 2018, 2019

2017, 2018, 2019

 

 

2020

2020

 

 

 

NPP Daily

 

 

NPP 5-day and Monthly composites

1996, 1997, 1998

 

 

 5-day all years

1999, 2000, 2001

 

 

 

2002, 2003, 2004

 

 

 

2005, 2006, 2007

 

 

NPP monthly;            NPP monthly PNG files

2008, 2009, 2010

 

 

 

2011, 2012, 2013

 

 

 

2014, 2015, 2016

 

 

 

2017, 2018, 2019

 

 

 

2020

 

 

 

 

References

Behrenfeld, M.J., Falkowski, P.G., 1997. Photosynthetic rates derived from satellite based chlorophyll concentration. Limnol. Oceanogr. 42, 1-20. https://doi.org/10.4319/lo.1997.42.1.0001

Kahru, M., Jacox, M.G., Ohman, M.D., 2018. CCE1: Decrease in the frequency of oceanic fronts and surface chlorophyll concentration in the California Current System during the 2014-2016 northeast Pacific warm anomalies. Deep Sea Res. Part I Oceanogr. Res. Pap. https://doi.org/10.1016/j.dsr.2018.04.007

Kahru, M., Kudela, R., Manzano-Sarabia, M., Mitchell, B.G., 2009. Trends in primary production in the California Current detected with satellite data. J. Geophys. Res. Ocean. 114, 1-7. https://doi.org/10.1029/2008JC004979

Kahru, M., Kudela, R.M., Anderson, C.R., Mitchell, B.G., 2015. Optimized merger of ocean chlorophyll algorithms of MODIS-Aqua and VIIRS. IEEE Geosci. Remote Sens. Lett. 12, 2282-2285. https://doi.org/10.1109/LGRS.2015.2470250

Kahru, M., Z. Lee, B.G. Mitchell and C.D. Nevison (2016), Effects of sea ice cover on satellite-detected primary production in the Arctic Ocean, Biology Letters, 12: 20160223. http://dx.doi.org/10.1098/rsbl.2016.0223.

Kahru, M., Kudela, R.M., Manzano-Sarabia, M., Greg Mitchell, B., 2012. Trends in the surface chlorophyll of the California Current: Merging data from multiple ocean color satellites. Deep. Res. Part II Top. Stud. Oceanogr. 77-80, 89-98. https://doi.org/10.1016/j.dsr2.2012.04.007

Reynolds, R.W., Smith, T.M., Liu, C., Chelton, D.B., Casey, K.S., Schlax, G., 2007. Daily high-resolution blended analyses for sea surface temperature. J. Clim. 20, 5473-5496, https://doi.org/10.1175/2007JCLI1824.1

Sathyendranath, S., Brewin, R., Brockmann, C., Brotas, V., Calton, B., Chuprin, A., . . ,Platt, T. (2019). An ocean-colour time series for use in climate studies: The experience of the ocean-colour climate change initiative (OC-CCI). Sensors, 19(19), 4285. doi: 10.3390/s19194285

Zhang et al. (2009) are used.