Processing MODIS satellite images for chlorophyll estimates against

Processing MODIS satellite images for chlorophyll estimates against automated fluorometer records Seppo Kaitala, Henrik Stutz Przemysław Bojczuk Finnish Institute of Marine Research Helsinki, Finland Seppo. Kaitala@Fimr. Fi Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Alg@line Flow-through recording and water sampling points by M/S Finnpartner • 24 sampling points • Every second week • 5 m depth • Flow through measurements with 250 m resolution Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Unattended sampling system Thermo salinograph Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

In situ chlorophyll-a fluorescence 31. 7 -2. 8. 2006 Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

MODIS AQUA Cloud mask 1. 8. 2006 1150 and in situ measurements Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Satellite image analysis Modis Aqua data (NASA GES Distributed Active Archive Center (DAAC ) Data Pool) for each 13 bands were exracted with HDFLook-Modis software and masking with Modis Aqua cloud masks were analyzed together with chlorophyll-a data with GRASS 5. 4 and 6. 1. Statistical analysis was done with PLS and PCR package in the R statistical software. Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

PLS validation Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Method: SIMPLS Number of latent variables considered: 1 -13 Suggested number of latent variables: 10 (cross validation) VALIDATION: Y RMS sd(RMS) 1 LV's 1. 0049 0. 07271 2 LV's 0. 7664 0. 05545 3 LV's 0. 7216 0. 05221 4 LV's 0. 6065 0. 04389 5 LV's 0. 5130 0. 03712 6 LV's 0. 4460 0. 03227 7 LV's 0. 4086 0. 02957 8 LV's 0. 4193 0. 03033 9 LV's 0. 3960 0. 02864 10 LV's 0. 3672 0. 02657 11 LV's 0. 3670 0. 02655 12 LV's 0. 3689 0. 02669 13 LV's 0. 3687 0. 02668 Q^2 0. 4428 0. 6755 0. 7122 0. 7970 0. 8547 0. 8902 0. 9077 0. 9031 0. 9134 0. 9255 0. 9256 0. 9248 0. 9249 Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne TRAINING: % variance explained X Y 1 LV's 72. 83 46. 80 2 LV's 98. 04 68. 88 3 LV's 99. 70 72. 52 4 LV's 99. 81 81. 27 5 LV's 99. 92 87. 06 6 LV's 99. 98 89. 79 7 LV's 99. 99 91. 61 8 LV's 99. 99 92. 56 9 LV's 100. 00 93. 30 10 LV's 100. 00 93. 75 11 LV's 100. 00 93. 83 12 LV's 100. 00 93. 88 13 LV's 100. 00 93. 88

Chlorophyll-a distribution estimation for 6 August 2006 Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Weekly averages August 2006, GRASS r. series Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Weekly maximum August 2006, GRASS r. series Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Baltic. Sea. Portal. FI Mapserver Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Mapserver • Geographic image maps and GIS data • Uses Shapelib, Free. Type, Proj. 4, GDAL/ORG. . . • University of Minnesota (UMN) – For. Net project, cooperation with NASA • Terra. SIP

Baltic. Sea. Portal ● ● Backend for Grass_Scripts Works on top of Mapserver Raster Query Map history Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Baltic. Portal Grass Scripts Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Mapserver, User interface; start at http: //itameripeli: 8080/ Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Mapserver, User interface Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne

Thank You for your attention Baltic Sea Portal prototype is founf at: http: //itameripeli: 8080/ Seppo. Kaitala@FIMR. fi FOSS 4 G 2006, 12 -15. Sept. 2006, Lausanne