www.ijrsa.org
International Journal of Remote Sensing Applications Volume 4 Issue 1, March 2014 doi: 10.14355/ijrsa.2014.0401.03
Monitoring Water Pollution of Lake Maryout on the Mediterranean Coast of Egypt 1 1 2 *
Samah Ahmed, 2*Mona F. Kaiser
Physics Department, Faculty of Education, Suez Canal University, El Arish, Egypt Geology Department, Faculty of Science, Suez Canal University, Ismailia, Egypt
[email protected]
Abstract The present study assessed serious eutrophication and pollution levels in Lake Mayout, Egypt. Chlorophyll spatial distribution and corresponding water surface temperatures in polluted water were extracted from 1986-2012 satellite images. Results showed a significant positive correlation coefficient (R2 = 0.99) between thermal water and chlorophyll abundance. Five classes were differentiated, including cultivated land, algal blooms, urban areas, and sea water from the two dimensional scatterplot between chlorophyll abundance and surface temperatures. Markedly high temperatures and very high NDVI values characterized cultivated land, while clean water exhibited low temperatures, and low NDVI. High surface temperatures and low NDVI were observed in planned urban areas. Finally, high surface temperatures and NDVI values characterized algal bloom polluted waters. Results showed surface areas exhibiting algal bloom were 29.2 km2 and water surface temperatures were 22ᵒC in 1986, and respectively increased to 32.5 km2 and 24ᵒC in 2012.
due to reclamation projects to convert the area to agriculture (El Rayis et al., 1994). Widespread sediments and a decrease in the water column, combined with continuous and massive inputs of untreated urban and industrial sewage, and agricultural pollutants have resulted in serious ecological degradation of Lake Maryout (El Rayis et al., 1994).
Keywords Lake Maryout; Water Surface Temperatures; Algal Bloom; Water Pollution; and Remote Sensing
Introduction Lake Maryout is the smallest of four coastal lakes of the Nile Delta. It forms the southern boundary of Alexandria City, along the Mediterranean coast of Egypt (Lat 31º07'N Long 29º57' E). The lake has an average depth of 90-150 cm (Alamim, 2009), with a nutrient range from 4.00 to 10.7 mg/l. The mean concentrations of all nutrients in Lake Maryout, except silicates, are several times higher than those in the other Nile Delta lakes. Low silicate content is due to the isolation from Nile water, which feeds other lakes with silicates (Saad et. al., 1984). Lake Maryout is a brackish water lake, which differs from other Delta lakes. It is a closed basin isolated from the open sea, with no natural connection to the Mediterranean Sea. During the last four decades, its area decreased from 700 km2 in 1818 to 60 km2 in 2012 36
FIG. 1 AREA OF STUDY
The lake is divided into four basins (Saad, 2003), including the Main Basin (A), Southwestern Basin (B), Northwestern Basin (C), and Fishery Basin (D) (Fig. 1). The lake serves as a drainage basin for the adjacent cultivated land, and the industrial and municipal wastes of Alexandria (Abdel Aziz and Aboul Ezz, 2004). Most landed fish come from its Main Basin. The water is suffering from pollution due to receiving untreated domestic sewage and industrial wastewater from Alexandria City (El Rayis, 2005). Since 1993, the Main Basin, receives flow from three main sources, which include two treatment plants: (1) Qalaa (agricultural), a drainage system that also carries discharge from the East Treatment Plant (ETP); (2) the West Treatment Plant (WTP) outlet; and (3) Omoum (agricultural) drain. The other three basins receive
International Journal of Remote Sensing Applications Volume 4 Issue 1, March 2014
flow from a variety of canals and drainage systems, but do not receive any effluent from ETP or WTP (UNEP, 2005). Before 1993, the lake Main Basin received most of the raw industrial wastewater and domestic sewage of Alexandria City, without treatment, as reported by Egyptian Environmental Affairs Agency (EEAA, 2009). Water surface temperature (WST) is an important parameter to examine the attributes characterizing aquatic ecology. WSTs in Lake Maryout exhibit a notable relationship with algal blooms. Miller and Rango (1984) reported a positive correlation between emitted thermal energy and algal concentration. Thermal infrared remote sensing is an effective tool to detect WSTs. Shaoqi et al. (2011) derived WSTs from Lake Taihu using a generalized single channel algorithm. WSTs vary among different areas, due to the direction and velocity of the water current. In addition, WST is typically higher in the areas surrounding the lake center. This may result from inland climate effect. The objectives of this study were to elucidate the relationships between phytoplankton and marine water conditions. Satellite data (i.e. remotely sensed data) and image processing techniques were utilized to retrieve water surface temperatures and algal bloom (chlorophyll) abundance in Lake Maryout. Data Sets and Methodology Image classification was applied to automatically categorize all pixels in an image into land cover classes or themes. Unsupervised classification was conducted using a histogram peak cluster technique to identify dense areas or frequently occurring pixels (Lillesand and Kiefer, 1994). In the unsupervised approach, spectrally separable classes are determined and defined relative to class informational utility to form a supervised classification scheme. Each class was verified in the field using a Garmin 38 GPS unit; more than 28 ground data sites were visited and confirmed. Efficient and accurate algorithms have been developed to retrieve sea surface temperatures with coarse spatial resolution instruments. However, a 1 km spatial resolution is not sufficient for water bodies less than 3 km in width, or less than 2 km from the shoreline. Therefore, higher spatial resolution instruments, such as the Landsat-5 TM or Landsat-7 ETM+ (Wloczyk, et al., 2006) were used. The availability of high-resolution remote sensing data has enabled resource managers to monitor land, coastal, and water resources at local
www.ijrsa.org
scales, thereby producing timely and reliable assessments. Digital image processing involves computer-aided manipulation and interpretation of digital images (Lillesand and Kiefer, 1994). Remote sensing and image processing are powerful tools for many applications and fields of research (Jeffrey and John, 1990). In the present study, image processing was conducted for Landsat images acquired for the 1986, 1990, 1999 and 2012 spring seasons. The thermal infrared bands from ETM+ images were converted to effective at-satellite temperatures using the following steps: 1. Transformation of Digital Number (DN) values to radiance ENVI 4.8 software was used to calculate radiance (W m-2 µm-1 sr-1) from (DN) values in the satellite images (bands 6, 10.4 to 12.5 µm) by the application of the Landsat calibration equation (Chander and Markham, 2003). Radiance values were modified by upwelling irradiance between the ground and the sensor (Fisher and J. Mustard, 2004). 2. At-satellite temperature transformation Sensor radiant is a function of surface temperature and emissivity. In Landsat images, band 6 captures the radiant thermal energy between 10.4 and 12.5 µm. The radiometric calibration method developed by the U.S. Geologic Survey (USGS) was utilized to transform the thermal infrared bands from ETM+ images to effective at-satellite temperatures (Schott et al., 2001, Chander and. Markham, 2003, Wloczyk et al., 2006 and Kaiser and Ahmed, 2013). Specific calibration coefficients and other related parameters of Landsat ETM+ images, including sun elevation gain and offset, were obtained from a level 1 product header or ancillary data record. TABLE 1 TM THERMAL BAND CALIBRATION CONSTANTS
The thermal infrared band was converted from spectral radiance to at-satellite temperature (Chander and Markham, 2003) using the following equation: K2 T= K ln( 1 + 1) Lλ where T is the effective at-satellite temperature (in kelvin), K2 is calibration constant 2 (in kelvin), K1 is the 37
www.ijrsa.org
International Journal of Remote Sensing Applications Volume 4 Issue 1, March 2014
calibration constant 1 [in W/(m2.sr.µm)], and Lλ is the spectral radiance at the sensor's aperture. Satellite temperatures were transformed from Kelvin to degrees Celsius. Table 1 shows the values of these parameters.
for 2012. All images were collected in April (spring season). ENVI 4.8 was processed to analyze remote sensing data. Unsupervised/Supervised classification (Isodata) was applied to monitor land cover changes during the last three decades (1986-2012).
3. Transformation of at-satellite temperature (in kelvin) to degrees Celsius
Five features were notable at the study site, (Fig. 2) including sea water (green), cultivated land (cyan), algal blooms (yellow), polluted water (blue), and urban areas (magenta) (Fig. 2). Phytoplankton blooms have a harmful effect on human health following human consumption of contaminated seafood (Wang et al., 2008). An increase in sea surface temperature is the catalyst for the occurrence of algal blooms. Rapid increases in algal cells during blooms may increase absorption of radiation, and consequently enhance the heating rate at the sea surface.
In this step, the final temperature was converted to degrees Celsius by subtracting 273 from the result obtained in step 2. The algal bloom (Phytoplankton) abundance was estimated using Normalized Difference Vegetation Index (NDVI). High algal bloom concentrations near several river outlets were due to chlorophyll-a abundance. Normalized Difference Vegetation Index (NDVI) was used to transform multispectral data into a single image band applying the standard algorithm equation derived by Jensen (1986). NDVI values indicated the amount of green vegetation in the pixel. Higher NDVI values indicated increased green vegetation; acceptable results were within the -1 and +1 range.
Algal bloom areas can absorb more solar radiation and heat the surface water temperature during the day (Wang and Tang, 2010). Algal bloom concentration for Lake Maryout was estimated by computing NDVI using Landsat images of red and near infrared bands (Shaoqi et al., 2011), (Fig. 3). In addition, Landsat thermal bands were utilized to calculate Lake Maryout water surface temperatures during the same time span (Fig. 4).
FIG. 2 MONITORING LAND COVER CHANGES DURING 1986-2012 IN LAKE MARYOUT
Results and Discussion Landsat image data used in this work included TM4 for year 1986, TM5 for 1990, ETM+ for 1999, and ETM+ 38
FIG. 3 MONITORING NDVI VALUES DURING 1986-2012 IN LAKE MARYOUT
International Journal of Remote Sensing Applications Volume 4 Issue 1, March 2014
www.ijrsa.org
water surface temperatures in polluted water within Lake Maryout were extracted from 1986-2012 satellite images. In addition, algal bloom surface area and WSTs increased from 29.2 km2 and 22ᵒC in 1986 to 32.5 km2 and 24ᵒC in 2012, respectively. NDVI and water surface temperatures showed a significant correlation coefficient (R2 = 0.99) (Fig. 5).
FIG. 4 MONITORING WST IN ᵒC, DURING 1986-2012 IN LAKE MARYOUT
Thermal classification was conducted using thermal infrared (10.4 to 12.5 um). This band region represents the infrared radiant flux amount emitted from the water surface. The sensor radiant is a function of surface temperature and emissivity, but is modified by upwelling irradiance between the ground and the sensor (Wloczyk et al., 2006). Previously reported radiometric calibration procedure (UNEP, 2005, Chander and Markham, 2003 and Fisher and Mustard, 2004) was provided to convert digital DNs. The thermal infrared band was converted from spectral radiance to satellite temperature. Results discriminated four classes, including cultivated land, algal bloom (polluted) water, urban areas, and sea water from the two dimensional scatterplot between chlorophyll and surface temperatures (Fig. 5). Lake Maryout (I) cultivated land was characterized by high temperatures, and very high NDVI; (II) sea water exhibited low temperatures, and low NDVI; (III) notably high surface temperatures, and low NDVI were observed in the urban planned areas; and (IV) the algal bloom (polluted) water showed high surface temperatures, and high NDVI values. Spatial distribution of chlorophyll and corresponding
FIG. 5 TWO DIMENSIONS SCATTER PLOT BETWEEN CHLOROPHYLL-a CONCENTRATION AND SURFACE WATER TEMPERATURE IN ᵒC
Conclusion Phytoplankton blooms result in harmful effects on human health following human consumption of contaminated seafood. A rise in sea surface temperatures is integral in the occurrence of algal blooms. Lake Maryout suffers from pollution due to untreated domestic sewage and industrial wastewater from Alexandria City. Landsat Thematic mapper images acquired from 1986, 1990, 1999, and 2012 were used to analyze and monitor algal (Phytoplankton) bloom concentrations in the lake. A significant positive correlation (R2 = 0.99) between thermal water and algal concentrations resulting from domestic sewage and wastewater effluent was detected. Chlorophyll abundance was estimated using NDVI values. DN values derived from thermal infrared bands were transferred from spectral radiance to atsatellite temperature to monitor water surface temperatures. From 1986-2012, water surface temperatures and pollution levels notably changed at Lake Maryout; algal bloom surface area occupied 29.2 km2 and water surface temperatures were 22ᵒC in 1986, and respectively increased to 32.5 km2 and 24ᵒC in 2012.
39
www.ijrsa.org
International Journal of Remote Sensing Applications Volume 4 Issue 1, March 2014
REFERENCES
p.750.
Abdel Aziz, N.E. and Aboul Ezz, S.M. "The structure of Zooplankton community in lake Maryout, Alexandria, Egypt." Egyptian Journal of Aquatic research ISSN 11100354, 30 (A), (2004): 160-170. Alamim "Alexandria Lake Maryout integrated management", EC-SMAP III, Final report, (2009): URL: http://www. medcities.org/ Chander, G. and Markham, B. "Revised Landsat-5 TM radiometric calibration procedures and post calibration dynamic ranges. IEEE Transaction on Geosciences and Remote Sensing," 41 (11), (2003): 2674-2677. Egyptian Environmental Affairs Agency, EEAA "Alexandria Integrated
Coastal
Zone
Management
Project",
Environmental and Social Impact Assessment. Ministry of State for Environmental Affairs (2009). El Rayis, O.A. "Impact of man`s activities on a closed fishinglake, Lake Maryout in Egypt, as a case study." Mitigation and Adaptation Strategies for Global Change, Springer, 10, (2005): 145-157. El Rayis, O.A., El Sabrouti, M.A. and Hanafy, H.M. "Some hydrochemical observation from Lake Mariut prior to diversion of sewage of eastern districts of Alexandria." Protec. (AMTA), (1994): 205-219. Fisher, J.I. and Mustard, J. "High spatial resolution sea surface climatology from Landsat thermal infrared data". Remote Sensing of Environment, 90, (2004): 293-307. Jeffrey, S. and John, E. "Geographic Information Systems, an introduction", Prentice-Hall, Inc., (1990): 303pp. Jensen,
J.R.
"Introductory
Digital
Image Processing",
Prentice-Hall, New Jersey, (1986): 379p. Kaiser, M.F. and Ahmed, S. "Optimal thermal water locations along the Gulf of Suez coastal zones, Egypt." Renewable Energy, 55, (2013): 374-379. Lillesand, T.M. and Kiefer, R.W. "Remote sensing and image interpretation." New York: Jon Wiley & Sons; (1994):
40
Miller, W. and Rango, A. "Using heat capacity mapping mission (HCMM) data to assess water quality", Water Resources Bulletin, 20, (1984): 493-501. Saad, M.H. "Impact of diffuse pollution on the socioEconomic development opportunities in the coastal Nile delta Lakes." Diffuse Pollution Conference Dublin, (2003): 6-81-6-85. Saad, M.H., El Rayis, O.A. and Ahdy, H.H. "Status of nutrients in Lake Maruit, a delta lake in Egypt suffering from intensive pollution." Marine Pollution Bulletin, 15(11), (1984): 408-411. Schott, J.R., Barsi, J.A., Nordgren, B.L., Raqueno, N.G. and Alwis, D. "Calibration of landsat thermal data and application to water resource studies." Remote Sensing of Environment, 78, (2001): 108-117. Shaoqi, G., Devong, S., Qiaohua, Z., Yunmei Li, Jiazhu, H. and Guokai, D. "Application of HJ/IRS to retrieve water surface temperature in Lake Taihu, China." International Conference on Information Science and Technology (ICIST), IEEE Xplore, (2011): 560-565. United Nations Environment Programme, UNEP. "Africa Environment Outlook2" (2005) . Wang, S.F. Tang, D.L. He, F.L., Fukuyo, Y. and Azanza, R. "Occurrences of Harmful algal blooms (HABs) associated with ocean environments in the South China Sea." Hydrobiologia, 596, (2008): 79-93. Wang, S.F. and Tang, D.L. "Remote sensing of day/night sea surface temperature difference related to phytoplankton blooms." International Journal of Remote Sensing, 31 (1718), (2010): 4569-4578. Wloczyk, C., Richter, R., Borg, E. and Neubert, W. "Sea and lake surface temperature retrieval from Landsat thermal data in Northern Germany". International Journal of Remote Sensing, 27 (12), (2006): 2489-2502.
