Chlorophyll-a Concentration Estimation And Its Diurnal Variation Obsevation Of Taihu Lake Based On GOCI

Chlorophyll-a (CHL-a) is a good indicator of alga species and concentration, of which the latter one is always deemed as critical parameter in assessment of eutrophic water. As there exists specific aborsorbing and scattering characteritic in CHL-a, optical remote sensing signal can identify water’s color and then quantify CHL-a concentration in water. During cyanobacteria’s eruption, CHL-a concentration dramatically changed within a day even in several hours. However, traditional remote sensing image’s temporal resolution is about one day or more, which cannot caputure cyanobacteria’s different developing stage continuously. GOCI, the first geostationary ocean color satellite in the world, offering8images of data whose temporal resolution achieve as short as1hour, ranging from8:25to15:25in Beijing time. Day May6,2012’s GOCI image was used in this paper to systematically monitor and analyze CHL-a concentration’s diurnal variation in Taihu Lake which would be of great significance to the emegency response and early-warning to eruption of cyanobacteria as well as severing great importance in offering suggestion to the harness of cyanbacteria in Taihu Lake.Based on field observation spectral reflectance and corresponding sample CHL-a concentration of different seasons from Oct.2010to May2012, this paper computed the correlationship between lake water spectral reflectance and CHL-a concentration. Result found several CHL-a concentration related bands including550nm,680nm and700nm. Semi-Empirical and Semi-Analysis models were tested, model precision and its application were evaluated, which indicated that Band Ratio model, First Derivation model and Three Bands model could achieve good estimation accuracy in Taihu Lake’s water CHL-a concentration; Three Bands Model which based on water radiactive transfer therory had a better estimation precision than Semi-Empirical Models; Three bands Model and Band Ratio Model are applicable to remote sensing image CHL-a inversion.Sun-Photometer observation data were used to precisely correct GOCI’s aerosol radiation. Taihu Station’s Sun-Photometer measurements and the AOT products were combined to compute the aerosol parameters and also the aerosol radiance. After eliminate all other factors such as Rayleigh scattering, white cap and sun glitter, atmosphere corrected GOCI remote sensing reflectance data were retrieved. Compared with in situ mesurements, results showed accepted precision. In visible band from443nm to680nm, the relative error are below10%; in Blue band of412nm, relative error is19%, relatively close to the required15%in world case II water ocean color water leaving radiance standard.To inverse CHL-a concentration of day’s Taihu Lake based on GOCI, models applicable to image inversion must be used. Thus, the Ratio model of bands660nm and680nm were used to estimate CHL-a concentration of GOCI’s8images of May6,2012, modeled precision up to80.1%. According to the day’s inversion result, we found that Taihu Lake’s CHL-a concentration was in a relatively high level in8:25, then from8:25to9:25, the CHL-a concentration was in the growth phase; from10:25, the CHL-a concentration began to decline gradually,15:25met the minimum level. This result fits well with the alga vertical migrations theory which insisted that when exposure to photosynthesis, alga buoyancy was controlled by cells bubble.