Top Navigation Bar

ASAE Conference Proceeding

This is not a peer-reviewed article.

EFFECT OF SUSPENDED SEDIMENT DISTRIBUTION ON SPECTRAL REFLECTANCE

V. Garg, and I. Chaubey

Pp. 136-140 in Total Maximum Daily Load (TMDL) Environmental Regulations II, Conference Proceedings, 8-12 November 2003 (Albuquerque, New Mexico, USA), ed. Ali Saleh. ,8 November 2003 . ASAE Pub #701P1503

Abstract

Remote sensing techniques have been used extensively to estimate optically active water quality parameters. Suspended sediment (SS) is the most common type of pollutant both in terms of weight and volume in inland waters. SS are helpful in determining water dynamics and spread of other pollutants. Laboratory studies done in past have developed regression between reflectance and uniform concentration of SS in water tank. However, depth distribution of SS is not uniform in inland waters and therefore, algorithms developed may be of limited applicability and accuracy. In this study reflectance was measured with time as SS settled in a water tank-giving rise to variation in SS concentration along the depth. A dual sensor Spectro-radiometer was used to measure relative reflectance in the electromagnetic spectrum region of 346 nm to 1000 nm (456 channel), with bandwidth of approximately of 1.438 nm. Reflectance values significantly changed with time even though overall SS volume remained same in water tank. It suggests that same SS volume with different depth distribution can give different reflectance values. Higher variation in reflectance was observed near 403, 576, and 807 nm spectral regions. Analysis is currently underway to regress the reflectance values with the volume of SS within the penetration depth of spectral region of reflectance. This analysis may provide us methodology to find SS volume in surface layer of water body.


INTRODUCTION

Suspended sediments (SS) are the most common pollutants both in terms of weight and volume in surface waters. Surface water quality degradation from suspended sediment loads derived from upland run-off has been identified as a major global problem (Brown, 1984; Walling and Webb, 1983). SS may serve as a surrogate contaminant in agricultural watersheds since phosphorus, insecticides, and metal adhere to fine sediments particles.

SS increase the radiance emergent from surface waters in the visible and near infrared portion of electromagnetic spectrum (Ritchie and Schiebe, 2000). Other water constituent which affect the absorption and scattering of the incoming light are light harvestin g pigments chlorophyll-a (chl- a) and cyaophycocyanin (cpc), humus, floating and submerged macrophytes, point source pollutants, e.g. oil spills. This property has been used to develop method for monitoring and assessment of water quality using satellite imagery. The advantages of using the synoptic coverage provided by sensors on satellite platforms for water quality assessment, such as cost-effectiveness, timeliness, and the ability for quantitative comparison of numerous water bodies, are well documented (Ritchie et al. l987). Spectral data allows for semi empirical methods based on knowledge of the specific absorption characteristics of water constituents in connection with regression approach (Thiemann, 2002; Fraser, 1998; Dekker, 1993; Doeffer, 1992).

In these studies water quality parameters were measured at some depth in the water body and that were regressed with the remote sensing data. Basic assumption in these studies is that the water quality parameters are uniform along the depth and a unique reflectance exists for a particular water quality condition. However, water quality parameters vary along the depth in a water body. Theoretical investigation of Nanu and Robertson (1993) showed significant changes in the shape of the simulated spectral reflectance when the suspended sediment distribution in the water column, extending from the surface to the light penetrating depth varies. In the present study variability of spectral reflectance is being determined due to settlement of the SSin water.

MATERIALS AND METHODS

Experiment was conducted at the Arkansas Agricultural Research and Extension Center (36:05:46.8N, 94:10:28.5W NAD83, 1294 ft NAVD88), an agricultural station of the University of Arkansas. Data was collected on 28th June 2003 under clear sky. The time of data collection was 1.00-to 3.00 p.m. local day saving time.

A volume of 2080 l of water in a vinyl tank, made from a 170 cm diameter filled up to a height of 91 cm was used for experiment. The walls and bottom were painted black to eliminate extraneous internal reflectance.

A Spectron Engineering SE- 590 spectroradiometer was used to collect radiance upwelling from the water. This instrument records a continuous spectrum in 512 bands. The spectral range of the instrument is from 336.02 nm to 1070.74 nm, with bandwidth of approximately of 1.438 nm. For this study data from 350 to 1050 nm were used because of significant noise in the water signal at wavelength shorter than 350 nm and longer than 1050 nm. The sensor head was attached to a stand and positioned over the tank at a height of 30 cm from the water level. A nadir view angle was selected for use (Novo et al. 1989). The 8 0 optic resulted in an instantaneous field of view 4.2 cm diameter circle on water surface. A notebook was used to initiate spectroradiometer scanning and storing the data. To avoid saturation of the sensor and use its full range, the instrument software controlled exposure time automatically. Also dark current correction option was selected for improving the measurement accuracy. A Barium surface reference panel served as the calibration standard. Reflectance factor (R ( ? )) were calculated using the following equation:

R ( ? ) = L( ? )/ B( ? )

Where L( ? ) is the wavelength-specific target radiance and B( ? ) is the corresponding radiance form the BaSO 4 panel. Five replicate scan were taken for each reading and the mean of two was used for the analysis. Silty loam soil was used for SS because they were available near experimental area. Soil was dried, sieved and weighed. Three hundred gram soil was used which generated 144 mg l -1 of SS concentration in a volume of 2080 l of water. Tap water was used for the experiment. First reading was taken without sediments. Then the soil was mixed uniformly in the water and reading was taken which was considered as the reading at 0 time and thereafter readings were taken at regular intervals without disturbing the water. More experiment is proposed with different concentration of SS.

Results And Discussion

Reflectance of water only (Figure 1) decreases with the increase of wavelength from 400 nm onwards and values were less from 750 nm to 1050 nm. After adding silty loam soil for creating SS concentration of 144 mg l -1 reflectance values changed. Readings were taken just after mixing the soil (T-00) and thereafter at every interval of 5 minute (Figure 1) of settlement of SS. Reflectance values changed with the settlement of sediments.

Between 400 to 1000 nm bands of electromagnetic spectrum, water with sediments had the higher reflectance value then the water only. The highest value of reflectance occurred near 585 nm. At 585 nm wavelength the reflectance value increased as the time of settlement increased. This indicates that wavelength near 585 nm were able to penetrate the water column of 0.91 m depth and therefore the sediment settlement at the bottom of the tank increased the reflectance near wavelength of 585 nm as the time of settlement of the SS increases. However, just opposite pattern of reflectance was observed near the smaller peak at wavelength near 815 nm. The value of the reflectance decreased as the time of settlement increased. This may be due to lesser penetration of the wavelength near 815nm. Settlement of SS with time reduced the concentration of SS in the upper layer of the water and thus decreased the reflectance.

Experiment is continuing to take reflectance at other concentration of the SS. Regression analysis will be done between reflectance and SS concentration of different depth.

136-140tmdl_files/image1.gif

136-140tmdl_files/image2.gif

REFERENCES

Brown, L.R. 1984. The global loss of top soil. Journal of Soil and Water Conservation. 39, 162-165.

Dekker, A.G. 1993. Detection of optical water quality parameters for eutrophic waters by high resolution remote sensing. Ph.D. Thesis, Vrije Universiteit Amsterdam, The Netherland.

Doerffer, R. 1992. Imaging spectroscopy for detection of chlorophyll and suspended matter. In: F. Toselli, & J. Bodechtel (Eds.), Imaging spectroscopy: fundamentals and prospective applications. ECSC, EEC, EAEC (pp215-257) (Brussel).

Fraser, R.N. 1998. Multispectral remote sensing of turbidity among Nebraska Sand Hills lakes. International Journal of Remote Sensing, 19 (15), 3011-3016.

Han, L., D.C. Rundquist, L.L. Liu, R.N. Fraser, and J.F. Schalles. 1994. The spectral responses of algal chlorophyll in water with varying levels of suspended sediment. International Journal of Remote Sensing, 15(18), 3707-3718.

Nanu, L., and C. Robertson. 1993. The effect of suspended sediment depth distribution on coastal water spectral reflectance: theoretical simulation. International Journal of Remote Sensing, 14 (2), 225-239.

Novo, E.M. L.M., J.D. Hansom, and P.J. Curran. 1989. The effect of viewing geometry and wavelength on the relationship between reflectance and suspended sediment concentration. International Journal of Remote Sensing, 10, 1357-1372.

Ritchie, J.C., and C.M. Cooper.1987. Comparison of Landsat MSS pixel array sizes for estimating water quality. Photgrammetric Engineering and Remote Sensing, 53, 1549-1553.

Ritchie, J.C., and F.R. Schiebe. 2000. Water quality, pp. 287-303, 351-352. In: G.A. Schultz and E.T. Engman (eds.), Remote sensing in Hydrology and Water Management, Springer-Verlag, Berlin, Germany.

Tassan, S. 1998. A procedure to determine the particulate content of shallow water from Thematic Mapper data. International Journal of Remote Sensing, 19(3), 557-562.

Thiemann, S., and Kaufmann, H. 2002. Lake water quality monitoring using Hyperspectral airborne data- a semiemperical multisensor and multitemporal approach for the Mecklenburg Lake District, Germany. Remote Sensing of Environment 81: 228-237.

Walling, D.E., and B.W. Webb. 1983. Pattern of sediment yield. Background to Palaeohydrology: A Perspective, edited by K.Gregory (Cambridge: Wiley) pp 69-100.