Browse by Year

Assessment of Lake Water Quality and Quantity Using Satellite Remote Sensing

June 21, 2019
(*Please email us at info@jyem.org for more information or to get access to the full article.)


AbstractAssessment of both water quality and quantity pose a great challenge to those studying the effects of anthropogenic activities on bodies of water. Eutrophication created by the increased concentration of nutrients including nitrates and phosphates has been known to contribute to the development of both toxic algal blooms, which serve as limiting factors in the ecosystems of the water, rendering it useless for consumption.1,2 Another common development is the buildup of suspended sediments (SS/TSS), contributing to the anoxic conditions characterizing environmental hypoxia.3 Because current methods for the assessment of the presence of such issues rely upon tedious and costly methods, a timely and cost-efficient method is desirable for application to the practice.4  This research relies upon analysis of the inherent optical properties of chlorophyll and sedimentation present within the bodies of water in question, achieved through analysis of the reflectance values of the red and blue bands from Landsat satellite images of five bodies of water. 5 The analysis, performed using Geographic Information System ArcMap, allows for determination of the values that attest to changes in surface area, turbidity, and eutrophication. The trends in the data hold consistency with the natural occurrences surrounding the bodies of water associated with the three parameters outlined above, supporting usage of remote sensing for qualitative and quantitative analysis of water.


Introduction

Lakes are popular hosts of environmental problems as a result of anthropogenic activities. For the majority of these lakes, causes of these problems often involve sediment loading or nutrient enrichment, also known as eutrophication.1 Eutrophication is also the cause of algal bloom in water. Both eutrophication and algal bloom are a natural phenomenon, but human activities may accelerate them, which can cause harm in terrestrial ecosystems. In fact, eutrophication and harmful algal blooms are the leading source of impairment of water quality in many lakes around the world.2 Specifically, human-derived sources due to industrialization, urbanization, or agricultural wastes due to the amount of excess nutrient that these sources then load onto their local freshwater bodies. Anthropogenic activities change the amount of Nitrogen and Phosphate - both of which are nutrients essential to algal growth - present in water. For instance, sewage, agricultural, and household discharges often contain large quantities of P minerals.3 Harmful algal blooms may cause anoxic conditions, which is the depletion of oxygen in water. Such conditions are especially dominated by cyanobacteria, which is a blue alga that produces cyanotoxins and makes lake water toxic, causing wildlife deaths and seafood poisoning in humans.4 

Traditional methods to measure water quality parameters like algal blooms involves field surveying techniques while measuring suspended solids involves the filtration technique. Unlike the other methods, studies show that satellite remote sensing is more cost-effective, economic, and ideal for acquiring spatial data from lakes with large surface areas7 like the ones that will be investigated. For the purpose of this study, there are two

other water quality parameters measured, besides the quantity factor with surface area. One is the chlorophyll, which will indicate the severity of algal bloom, and the other is total suspended solids, as a measure of water turbidity. The Inherent Optical Property (IOP) - which refers to absorption and scattering properties of underwater contents - of chlorophyll and suspended solids were used to determine algal and sediment presence. And because of the optical properties of chlorophyll and suspended solids in water, one can use commercially available optical instruments to measure their respective concentrations.7 This can be applied to satellite data because of the way in which satellite sensors collects the intensity of light reflected. And since satellites measure reflectance values in different intervals of the electromagnetic spectrum, the focus will be placed on reflectance values on certain intervals - also known as band values - in this paper. In summary, a lower reflectance value of blue band correlates to a higher concentration of sediments. As a lower reflectance value of the red band would suggest a higher presence of chlorophyll.

In this study, three lakes across the world are analyzed, and each is chosen for the significance of their impact on local livelihood. The three bodies of water are the Aral Sea in Kazakhstan, the Wular Lake in Kashmir, and Lake Taihu, or Lake Tai in China.

Methods

United States Geological Survey satellite images were to collect the data related to the measures of surface area, turbidity, and eutrophication. The satellite images were acquired from the United States Geological Survey’s Earth Explorer Database. The GIS software was utilized in order to determine the surface area and mean red and blue band values for each of the bodies of water.6 The satellite images that were selected were without any cloud coverage over the bodies of water, as the functions that were utilized in determining the presence of chlorophyll and sedimentation relied upon the properties of reflectance of light.6 The mean red band values of each image would analyze the levels of chlorophyll present in the lake, while the mean blue band values would represent the presence of Total Suspended Solids (TSS). The selected images were then downloaded with the “ LandsatLook Images with Geographic Reference” option in order to be able to have the images automatically oriented geographically upon downloading, taking advantage of the automatic georeferencing done by ArcMap. Following the creation of a mosaic image, a new shapefile was created and categorized as a polygon, to permit usage of the editing tool that allows users to outline figures. The shapefile was edited to create features known as “profiles.” The “Freeform” tool was utilized in creating the profiles over the bodies of water, as it is designed to function as a tracer. This ability permitted the creation of accurate profiles that covered the bodies of water. The profiles were created with the intention to analyze to mean values for the red and blue bands of the lakes. In order to receive the intended values from only the areas that were covered by a profile, the “Clip” tool found under the Raster Library was utilized. This tool allows users to make a copy of the areas that are underneath the created profiles. Clipping the shapefile to the mosaiced image creates the copies of the profiles, that appear as a new layer on ArcMap. ArcMap automatically computes the data values that are associated with the new layer and attaches them to the newly created layer. To calculate the surface area, the attribute tables were opened. The area is not automatically calculated, but can be computed using the software. An attribute was added to attribute file: “Surface Area, adding the values of the surface areas of the shapefiles created. For each clipped layer, the properties feature was utilized to access the statistical analysis section, locating the values listed as “mean.” The values were listed as Band 1, Band 2 , and Band 3.  Each pixel that forms an image derives its color from the values of intensity of three different bands: Red, Green, and Blue. Band 1 and Band 3 were the bands that were observed as they represent the values of the bands that attest to the degrees of turbidity and eutrophication: the conditions in question.

Data and Results

The data visualizations shown below display the conglomeration of the data that were collected using the methods outlined above. The graphs present the calculated mean red and blue band values as represented by the bars of the respective colors. The solid black line represents the surface area of the lake for that specific year.
In regards to the most drastic change in surface area, the Aral Sea suffered from the
greatest decrease in surface area during the two decade period in which the available data samples were analyzed. Below are two images that were used in the study: the first on the left is from the sample used to analyze the desired values for the year 1999. The photo on the right is the image that was analyzed to retrieve the data for the year 2017.

                                           Figure 1.  Data for the Lake Taihu

The figure below shows the Landsat­7 - image of the Aral Sea Aral Sea in the first year of the study: 1999. The figure below shows the Landsat­7 - image of the Aral Sea Aral Sea in the final year of the study: 2017

                                                                                                   Figure 2. Aral Sea in 1999 and 2017

The two figures show the Aral Sea in 1999 and 2017 respectively.

 

                            Figure 3. Graph of Aral Sea with all three parameters

In Figure 3, the blue represents blue band values, red represents red band values, and black line shows the trend for surface area 

                               Figure 4. Images of Lake Taihu in 2001

The figure above shows an image of Lake Taihu in 2001, maintaining non uniformity about it in its color, with the northern region of the lake suffering from greater nutrient concentration, creating a faint green tint.

 

                                 Figure 5. Images of Lake Taihu in 2016

The figure above shows an image of Lake Taihu in 2016, having maintained a more consistent and clearer body of water, attesting to the decrease in pollution.

The figures above show an images of Lake Taihu in 2001 and 2016, maintaining non uniformity about it in its color, with the northern region of the lake suffering from greater nutrient concentration, creating a faint green tint, and then having maintained a more consistent and clear body of wate in 2016, attesting to the decrease in pollution.

 

                       Figure 6. Graph of Wular Lake with all three parameters

Blue representing blue band values, red representing red band values, and black line shows the trend for surface area.)

Discussion and Conclusion

From observing the images of the Aral Sea, there is an apparent decrease in surface area from the almost two decade period that was examined. This is greatly reflected in the rapid decrease in value of the surface area as represented on the graph. In addition, there is a trend present in the values of the red and blue bands where the values are present: the values of the red bands are significantly higher than the values for the blue band means, though this disparity does seem to decrease in the more recent years. The decrease in the red band value from the year 1999 to the 2017 reflects the presence of organisms that absorb more red light, representative of the occurrence of algal blooms that are caused by the phenomenon of eutrophication. The minor increase in blue bands may allow us to determine a decrease in the presence of total suspended solids present within the lake.

For the Wular Lake, the values for the blue bands are consistently significantly higher than the values for the red bands, slowly decreasing in recent years until meeting values similar to those of the red bands. The Wular Lake shows to have maintained a consistency with the values of the red bands as they do not display a significant gap, showing a stability of the values representing the presence of chlorophyll in the lake. There was a significant decrease in the values for the blue bands, representing an overall increase in the quantity of total suspended solids in the lake. The surface area for the lake remained rather stagnant with the exception of 2014, during which floodwaters increased the surface area of the lake. It otherwise did not exhibit any significant change during the years that were utilized in the data extracted.8

 With both of its red and blue band values observed to have an increasing trend, Lake Taihu is the only lake in which its case with pollution is gradually getting better over the course of the 15-year period. The increase in band values is more notable as the mean of the reflectance value in the blue wavelength is nearly three times at 2016 than it was 2001. This suggests a significant decrease in the presence of TSS. The increase in both band values also suggests a decrease in chlorophyll presence in the lake. The surface area of Lake Taihu has remained relatively stagnant over the 15-year period.

It is relevant here to mention that not all of these lakes were expected to have large fluctuations in all three quality as well as quantity parameters to begin with. For instance, during the lake selections phase, it is expected for lakes like the Aral Sea to show a more obvious trend in decreasing surface area for the past few years, because it is more notoriously known globally for its problem with the shrinking size. Other lakes, like Lake Tai and Wular Lake, are not expected to have as much of a decrease in surface area, though it is expected to have more problems in terms of its water quality, as they are often subject to case studies involving the extents of their harmful algal bloom or excessive sedimentation.

Looking at the red band values of these graphs, there seems to be an observable trend in all the lakes except for the Aral Sea. Which makes sense because the Aral Sea is the only lake out of all the five that is technically located in the middle of a desert in Central Asia, and it seems to be the most remote from live plants and vegetation. The rest of these lakes are most located in scenic areas where there are mountains full of trees some of them are located in a more subtropical climate. In the Wular Lake, for example, there is the most apparent trend of a decrease in red band means, which is correlated to an increase in chlorophyll concentration. This is an indicator that the algal growth in Wular Lake is certainly still an ongoing problem. Lake Tai, however, seems to be in the minority as there is an upward trend in the red band value, indicating a decrease in algal growth. This is evidently consistent with its local government’s conservation efforts to control local industrial pollutions. The blue band values seem to be fluctuating from year to year. The only cases with an obvious trend may have been present is in the case of Wular Lake. There is a relatively strong trend of decreasing blue band values, which indicates an increase in suspended sediments. This fits context as reportedly, the lake still suffers from pollutions from fertilizers and animal manure from plantations nearby. Another trend in blue band is observed in Lake Tai, as there is a slight increase in blue band values, meaning there has been a decrease amount of sediments.

References

[1] Smith, V., Tilman, G., & Nekola, J. (1999). Eutrophication:Impacts of excess nutrient inputs on freshwater, marine,andterrestrial ecosystems. ​Environmental Pollution,100(​ 1-3), 179-196. doi:10.1016/s0269-7491(99)00091-3 12     

[2] Chislock, M.F.; Doster, E.; Zitomer, R.A.; Wilson, A.E. (2013)."Eutrophication: Causes, Consequences, and Controls in Aquatic Ecosystems". Nature Education Knowledge. 4 (4): 10. Retrieved 10 March 2018.                                                            

[3] Anderson, D. M., Glibert, P. M., & Burkholder, J. M. (2002). “Harmful algal blooms and eutrophication: Nutrient sources, composition, and consequences.” Estuaries,25(4), 704-726. doi:10.1007/bf02804901 

[4] Bush et al. (2017). "Oxic-anoxic regime shifts mediated by feedbacks between biogeochemical processes and microbial community dynamics". nature. Bibcode:2017NatCo...8..789B. doi:10.1038/s41467-017-00912-x. 

[5] Michaud, Joy P. (1994). "Measuring Total Suspended Solids and Turbidity in lakes and streams." Archived 2010-07-30 at the Wayback Machine. A Citizen's Guide to Understanding and Monitoring Lakes and Streams. State of Washington, Department of Ecology.                                        

[6] Alesheikh, A. A., et al. “Coastline Change Detection Using Remote Sensing”. International Journal of Environmental Science & Technology, vol. 4, no. 1, Jan. 2007, pp. 61–66., doi:10.1007/bf03325962.                                                      

[7]Babin, M., Cullen, J., Roesler, C., Donaghay, P., Doucette, G., Kahru, M., . . . Sosik, H. (2005). New Approaches and Technologies for Observing Harmful Algal Blooms. Oceanography, 18(2), 210-227. doi:10.5670/oceanog.2005.55 [8] Stony, J., & Scaramuzza, P. (n.d.). Landsat 7 Scan Line Corrector-Off Gap-Filled Product Development.