River Water Quality Analysis of Hadano Basin and its Relationship with Nonpoint Sources of Pollution

ABSTRACT The water quality of six rivers in Hadano basin was investigated and its relationship with nonpoint sources of pollution was analyzed. This study was undertaken to spatially examine the present status of the river water quality of Hadano basin. Ground water circulation influenced both the water quality and quantity. In the downstream basins of Muro and Kuzuha Rivers, COD and TP were diluted by the ground water inflow. In Mizunashi River and the upstream of Kuzuha River, water infiltrated to the subsurface due to the higher permeability of the river bed. Chemical oxygen demand (COD), total phosphorus (TP) and total dissolved solids (TDS) showed good correlation with unsewered population and agriculture area. While total nitrogen (TN) had good correlation with atmospheric nitrogen (N) deposition loads. Multiple regression analysis between water quality pollution loads and influencing factors revealed that unsewered population had higher impact on the river water quality; nonetheless, agriculture also had some effects. For TN, atmospheric N deposition load was taking effect, implying that it plays a significant role on the water quality and cannot be denied for proper water quality management. Development of sewerage system could be considered as the decisive factor to maintain the river water quality in the Hadano basin.

River Water Quality Analysis of Hadano Basin and its Relationship with Nonpoint Sources of Pollution

Gaurav SHRESTHA*, Satoru SADOHARA*, Shigeki MASUNAGA*, Hiroaki KONDO**, Satoshi YOSHIDA*, Yuichi SATO*

*Graduate School of Environment and Information Sciences, Yokohama National University, 79-7 Tokiwadai, Hodogaya-ku Yokohama 240-8501, Japan

**Research Institute for Environmental Management Technology, National Institute of Advanced Industrial Science and Technology (AIST), 16-1 Onogawa, Tsukuba 305-8569, Japan

ABSTRACT

The water quality of six rivers in Hadano basin was investigated and its relationship with nonpoint sources of pollution was analyzed This study was undertaken to spatially examine the present status of the river water quality of Hadano basin Ground water circulation influenced both the water quality and quantity In the downstream basins of Muro and Kuzuha Rivers, COD and TP were diluted by the ground water inflow In Mizunashi River and the upstream of Kuzuha River, water infiltrated to the subsurface due to the higher permeability of the river bed Chemical oxygen demand (COD), total phosphorus (TP) and total dissolved solids (TDS) showed good correlation with unsewered population and agriculture area While total nitrogen (TN) had good correlation with atmospheric nitrogen (N) deposition loads Multiple regression analysis between water quality pollution loads and influencing factors revealed that unsewered population had higher impact on the river water quality; nonetheless, agriculture also had some effects For TN, atmospheric N deposition load was taking effect, implying that it plays a significant role on the water quality and cannot be denied for proper water quality management Development of sewerage system could be considered as the decisive factor to maintain the river

water quality in the Hadano basin

Keywords: atmospheric nitrogen, unsewered population, water quality

INTRODUCTION

Water is one of the most abundant elements found on Earth It forms an essential component of the ecosystem and its services (Millennium Ecosystem Assessment, 2005) However, because of modernization and urbanization, there is an increasing and concurrent prevalence of problems in water environment Degradation of vital water resources indicates the loss of natural systems and the services they provide (Carpenter

et al., 1998) Japan has a long history of struggle with water environment conservation and it is still endeavoring to achieve conservation (Otsuka et al., 2009) The rapid

urbanization of Japan over the last 50 years has caused changes in land use and lifestyle, which have affected the river water quality (Tabayashi and Yamamuro, 2009) UNESCO (2003) stated that in seven prefectures, including Tokyo, the amount of land used for building in 2000 had increased by 7.2% as compared to 1974, while agricultural land had reduced by the same percentage Increase of nitrogen (N) and phosphorus (P) loads to rivers due to land use alteration has been the major sources of water pollution (Oki and Yasuoka, 2008)

Remarkable water quality improvements have been observed over the recent years because of the reinforced regulations imposed on the industrial wastewater and the consequent development of sewerage systems However, pollution loads from

households and agricultural lands are still high (UNESCO, 2003) Besides, the coverage ratio of the sewerage system is still low in small and medium cities (Kohata and Mizuochi, 2007)

Nonpoint source pollution such as runoff from unsewered developments and urban area runoff accompanied by agricultural runoff with higher amount of fertilizer, contribute to

the significant amount of P and N to surface waters (Carpenter et al., 1998) In general,

unsewered pollution is treated as a point source of pollution since it is caused by the domestic wastewater discharged from households When wastewater generated from unsewered areas as a whole is considered, then it is taken as nonpoint source of

pollution (Carpenter et al., 1998; Loague and Corwin, 2005) These pollutants generally

exist in medium to low concentrations but their behavior cause them to be widely distributed in nature, making their sources difficult to identify and this can be the

greatest threat to surface waters (Corwin et al., 1997; Verro et al., 2002; Li et al., 2004; Banadda et al., 2009) As the river basins in Japan are commonly characterized by

rugged hilly terrains, rivers are steep and short so pollution loads discharging from catchments are basically generated from nonpoint sources, scattered over various land covers (Oki and Yasuoka, 2008) Analyzing the impact of nonpoint sources of pollution

on water quality is difficult due mainly to the difficulty in their identification (Sharpley

et al., 2001; Bahar et al., 2008)

In the present era of urbanization, the effect of atmospheric N on water bodies cannot be neglected Atmospheric N is one of the complicated nonpoint sources that can seriously affect the river water environment because it has been known that reactive N is now accumulating in the environment on all spatial scales; local, regional and global

(Galloway et al., 2003) Recently, the deposition of atmospheric N has been gradually

increasing in the forested areas like Tanzawa Mountains of Hadano City where Kaname and Mizunashi Rivers originate It had been reported that the forested watersheds neighboring the Kanto District of Japan had the highest levels of nitrate export to

streams (Fujimaki et al., 2008), indicating that the atmospheric N generated from the

heavily urbanized Kanto region is presumably playing an important role in the N concentrations of river waters Generally, urban areas have higher atmospheric N pollution caused by anthropogenic activities as a consequence of energy production

Hadano City is known for its abundant water resources including surface water, ground water and springs However, there was a problem of water quality deterioration in the past and the problem still persists with the variation in pollution scale Rivers in Hadano City are polluted due to the inflow of domestic wastewater, urban and agricultural

runoffs (Fujino et al., 1997) Being a middle-sized city, Hadano City is comprised of

various types of land cover The core areas of the city are mostly urbanized, whereas rural areas with dense forest and agriculture are also present in uplands Hadano City has been experiencing a gradual rate of urbanization with corresponding expansion of newly developed areas, whereas the development of sewerage system is still behind the pace of urbanization With these perspectives, Hadano City as a research area is a spatial combination of different land uses such as developed and rural areas Most of all, Hadano City with its abundant ground water and existing water circulation is a very complex system regarding water quantity and its influence on water quality Subsurface infiltration of water and natural outflow of ground water (spring) are the salient features

of this city This study on the analysis of river water quality and its relationship with nonpoint sources of pollution including the complex ground water circulation and atmospheric N, that has not been elucidated yet, can be effectively employed for water quality management of Hadano basin as well as other watersheds, both with and without complex ground water circulation Therefore, as a research area of this study, it is significant and it has socio-economic values in terms of water quality management With the start of the sewerage system in 1978 and its progressive development, the river water quality in the Hadano basin became better with time Nevertheless, in the areas where installation of sewerage system is still in progress and where it is not introduced, improvements in water quality have not been observed yet and it has been a big problem hereafter Furthermore, the rapid rate of urbanization was also affecting the river water

quality in Hadano (Fujino et al., 1997) Therefore, nonpoint sources of pollution such as

non-development of sewerage system, urban area and agriculture, as well as atmospheric N deposition loads could be considered as determining factors of the water quality in this basin

Understanding the pollution scenario of the whole basin in a broader scale is necessary for the proper management of river water quality With this in mind, this study was undertaken to spatially investigate the present status of the river water quality in Hadano basin and its relation with nonpoint sources of pollution Up to now, river water quality had been monitored only at the downstream of rivers This monitoring system could not comprehend in detail, the water quality condition of the whole river system In addition, the study on the relationship of river water quality with sewerage system and land use has not been done yet in the Hadano basin Thus, with a holistic approach, this study was carried out to spatially examine the water quality variation of each river at its upstream, midstream and downstream, including that of tributaries and service water canals The inclusion of atmospheric N deposition load as one of the factors that influence water quality is completely new in this study and has not been considered before

METHODS

Study Area and Rivers

Hadano City lies in the western part of Kanagawa Prefecture with an area of 103.61 km2

In the north, Tanzawa mountain range is situated and Shibusawa Hill exists in the south Six rivers of Hadano City (Muro, Mizunashi, Kuzuha, Kaname, Ohne and Shijuhase rivers) were studied Mizunashi River is flowing at the central region and at its eastern part, Kuzuha and Kaname Rivers are flowing, forming an alluvial fan in the plains Muro River is flowing along the Shibusawa faults at the southern part Ohne River is flowing at the southeast part and Shijuhase River at the western region, flowing from north to south (Fig 1) Muro, Mizunashi, Kuzuha and Kaname Rivers merge at the southern end, then flows as Kaname River Kaname River, originating from Tanzawa

mountains, is the longest river of Hadano City with the length of 21 km (Fujino et al.,

1997)

Monitoring Stations and Delineation of Drainage Basins

Water quality monitoring stations were allocated at 32 locations (Table 1) along the

main rivers, tributaries and service water canals This was done depending upon the condition of sewerage system, land use pattern and local conditions The monitoring stations were allotted at upstream, midstream and downstream sections of rivers as well

as at the outlet of respective tributaries and service water canals inflowing to rivers

Drainage basins corresponding to each monitoring station were delineated (Fig 2) with the help of ArcGIS’s Spatial Analyst tool for hydrologic analysis, using a Digital Elevation Model (DEM) (Maidment, 2002; Kawasaki, 2006) Drainage basins were delineated estimating that water within each basin drains to their respective monitoring stations The prepared data were modified on the basis of the detailed rainwater

8

9 10 11 12 13

Fig 1 - Hadano Basin and its river system Fig 2 - Locations of water quality

monitoring stations Table 1 - Water quality monitoring stations and the corresponding river basins

No Monitoring Station River Basin No Monitoring Station River Basin

1 Chimura Muro 17 Shimo-ochiai (Tributary) Kaname

2 Ishiuchiba Muro 18 Irifuna (Tributary) Kaname

3 Neshitabashi Muro 19 Kaguchi (Tributary) Kaname

4 Sakagawa (Tributary) Muro 20 Shinsaibashi Kaname

5 Yamanokami Mizunashi 21 Saigabun (Service Water Canal) Kaname

6 Minasebashi Mizunashi 22 Neshita-yohsui (Service Water Canal) Kaname

7 Sakurabashi Mizunashi 23 Shimo-oduki (Service Water Canal) Kaname

9 Sakurazawabashi Kuzuha 25 Minamiyanazawa (Tributary) Ohne

10 Shiyamabashi Kuzuha 26 Jakubozawa (Tributary) Ohne

12 Kuzuhaohashi Kuzuha 28 Shingawa (Tributary) Ohne

15 Kanamegawabashi Kaname 31 Kawachibashi Shijuhase

16 Ochiaibashi Kaname 32 Yanagawa (Tributary) Shijuhase

drainage area map of Hadano City Then, to verify the accuracy of data with respect to the real condition of the study site, they were checked by the concerned officers of Hadano City and the data were revised based on their advices Accurate locations of the monitoring stations were digitally recorded by using Mobile GIS (ArcPad and GPS) (Fig 2) Ochiaibashi monitoring station of Kaname River was situated before any rivers merged with it and Shinsaibashi monitoring station was located after Kuzuha, Mizunashi and Muro rivers along with its other tributaries merged with it

Water Sampling and Chemical Analysis

At the allocated monitoring stations (Table 1), water samples were collected once every two months from May 2009 to March 2010 Water samples (1500 mL each) were collected manually at each station using polyethylene bottles (1000 mL and 500 mL) The pre-washed bottles were rinsed thrice with water samples on the site before sample collection Water samples were stored in a cooler box and transported to the laboratory

Chemical oxygen demand (COD) concentration was determined by analyzing the oxygen demand by potassium permanganate at 100oC (CODMn) (JIS K0102 17) Total nitrogen (TN) concentration was determined by ultraviolet absorption photometry (JIS K0102 45.2) and total phosphorus (TP) concentration was determined by potassium peroxydisulfate resolution method (JIS KO102 46.3.1) These analytical methods for the determination of COD, TN and TP concentrations were based on the testing methods for industrial wastewater, Japan Industrial Standard (JIS) KO102 published by Japanese Standards Association in 2009 Electrical conductivity (EC) was measured on-site using

an EC meter (ES-51; Horiba, Tokyo) The EC meter was first calibrated using a standard solution of potassium chloride A conversion factor was used to estimate total dissolved solids (TDS) (mg/L) from EC (µs/cm), which depends on the salts specifically present in the water In this study, the conversion factor of 0.7 was considered (Walton, 1989) At the same time, river flow velocity was also measured with a flow velocity meter (CM-1BN; Toho Dentan, Tokyo) at each station River flow was calculated by multiplying the river cross-sectional area by the flow velocity at various points along a transect across the rivers and tributaries

Sampling was always done in clear weather condition to prevent any abrupt changes in measurements, except in January which was influenced by an unpredicted rainfall To avoid unsteady conditions, sampling was not conducted within 3 to 4 days after rainfall events Sampling and measurements along each individual river were done continuously from upstream to downstream and at its tributary Only after this that the sampling was done in other rivers There were three working groups sampling in three different rivers

at the same time to shorten the time lag as much as possible The monitoring sequence was always in the order: Kaname, Kuzuha, Mizunashi and Muro rivers as they merge with one another Then only Ohne and Shijuhase rivers were monitored It took about five hours on the average to monitor six rivers in each monitoring day Thus, the average time consumed for the monitoring of each river was almost about an hour a day

Data Collection and Preparation

Data of areas where the sewerage system had been installed until 2009 were acquired from Hadano City as a sewerage system data in paper format These data were digitally prepared in ArcGIS (Fig 3) Population data were obtained from the national census

data of Japan and the data of each drainage basin were prepared in GIS and were calculated based on area proportion (Fig 4) Land use data of 2005 used in this study were acquired from Kanagawa Prefecture City Planning Basic Survey data Land use data consisted of 14 categories (Fig 5), which were recategorized into 4 categories: paddy field, cultivated land, forest and urban area The river, water body and seashore were not considered While recategorizing, open space, residence, park, business, industry, agriculture facility, road and railway were categorized as urban areas Similarly, abandoned farm was merged with cultivated land Paddy field and forest were used as it

is (Table 2)

Fig 5 - Land use Data

LandUse 2005 Open space Residen ce Park

Bu siness Forest Ind ustry River, waterbody Seashore Pad dy field

Cu ltivated land

Ab an doned farm Agricu lture facility Road

Railw ay

Fig 3 - Sewerage System Data

Sewerage System

Sewered Area Drainage Basin Hadano Basin

Fig 4 - Population Data

Population 2009 (Person)

Table 2 - Recategorization of Land use

Land Use Category Recategorized Areas

1 Paddy field Paddy field

Land use data of each drainage basin was extracted in GIS and the area of each land use type was calculated (Fig 6).Building data with floor area were also acquired from Kanagawa Prefecture City Planning Basic Survey data Then, unsewered population was calculated in GIS by overlaying layers of sewered areas, households with total floor areas and population data for each basin On the basis of sewered areas and household data, households without connection to the sewerage system were distinguished first (Fig 7) After that, unsewered population was calculated using unsewered household buildings and population data, based on area proportion (Fig 8)

Overlay of sewered area

Sewered Area Minasebashi Basin

Separation of unsewered households

Unsewered Household Sewered Household Minasebashi Basin

Overlay of sewered area and households

Household Sewered Area Minasebashi Basin

Fig 6 - Land use proportion of each drainage basin

Fig 7 - Separation of unsewered households of Minasebashi basin of Mizunashi River

Pollution Load of Individual Basin

As river flows downstream, pollution from upstream basins are also carried to downstream basins Thus, pollution observed at downstream basin also includes pollution loads from its upstream basin in addition to those contributed by the pollution sources within the basin However, pollutants headed downstream get self-purified to some extent because of physical processes like dilution, diffusion and settling; chemical processes like oxidation, reduction and adsorption; and biological processes like decomposition and uptake by organisms

Fig 9 shows the upstream Chimura, midstream Ishiuchiba and downstream Neshitabashi sub-basins of Muro River basin Pollution loads at Chimura sub-basin include the loads of this basin only While at Ishiuchiba sub-basin, pollution loads

Unsewered Population (Person)

Chimura

Ishiuchiba Neshitabashi

# Monitoring Station Chimur a

Ishiuchiba Neshitabashi River

Fig 8 - Unsewered population of each drainage basin

Fig 9 - Upstream, midstream and downstream basins of Muro River

▲

include those from upstream Chimura basin in addition to loads from this basin Similarly, at downstream Neshitabashi sub-basin, pollution loads are also contributed by the upstream basins Chimura and Ishiuchiba, in addition to its own loads In this study,

as the water quality was monitored at the upstream, midstream and downstream sections

of each river, analysis was done considering the water quality pollution load of individual basin (Pn) of each monitoring station, so that the pollution scenario and the related influencing factors of each individual basin could be analyzed in detail, reflecting its own characteristics This was calculated through equation (ii) below by deducting the pollution loads observed at the upstream basins (Pon-1) multiplied by pollution remnant rate (Rr) after self purification from those observed at downstream ones (Pon)

The general equation for pollution load is given as

1000

QC

………… (i)

where P: pollution load (kg/day)

C: concentration of water quality parameter (mg/L)

Q: river flow (m3/day)

while the pollution load of individual basin was calculated as (Yoshida and Yasui, 1992; Modified from Ministry of Construction, 1999)

Pn = Pon – Pon-1 × Rr ……… (ii)

where Pn: pollution load of nth basin

Pon: pollution load observed at the monitoring station of nth basin

Pon-1: pollution load observed at the upstream monitoring station of nth basin Rr: pollution remnant rate after self-purification

The pollution remnant rate after self-purification was calculated as

Rr = exp (-Kr • t) ………… (iii)

where Kr: self-purification coefficient (day-1)

t: time for pollution to flow from upstream to downstream station (day)

Meanwhile, t was calculated as

Lvt86.4

where L: river section length (km)

v: river flow velocity (m/s)

While calculating the pollution load of the individual basin, the influence of ground water circulation, such as infiltration and inflow of ground water were also taken into consideration on the basis of river flow survey In the rivers with highly permeable geology, water easily infiltrates to the subsurface and as a result river flow is less and in some sections water does not flow at all Consequently, pollution loads will not flow to

downstream basins In these cases, it was considered that upstream basins do not contribute to pollution in downstream basins On the other hand, in the drainage basins that experience the influence of ground water inflow, river flow is higher than its upstream, thereby diluting the pollution In these contexts, the dilution factor was calculated based on the monitored river flow and considered in the analysis for COD and TP In case of TN, its inorganic form (NO3-) gets easily transported to ground water

so ground water can contribute to N concentrations in the river water Likewise, it can also contribute to TDS concentrations Therefore, dilution factor was not considered for

TN and TDS Detailed explanation is presented in the results and discussion section

Self-Purification Coefficient (Kr)

With reference to the values of several rivers in Japan reported by Nagasawa and Teraguchi (1971), Yoshida and Yasui (1992), and MLIT (2003), self-purification coefficients (Kr) were selected from 0.5 to 2.5 with 0.5 interval, i.e 0.5, 1, 1.5, 2 and 2.5 Pollution loads of individual basins (Pn) were calculated for each Kr at different time periods of monitoring, using equation (ii) Then, Kr was chosen by comparing the calculated pollution loads of individual basins with the pollution loads generated from nonpoint sources of pollution within the respective basins such as unsewered population and different land uses, which were calculated by using the unit loads of pollution reported by the Ministry of Construction, Sewerage and Wastewater Management Department (Ministry of Construction, 1999) (Table 3) Detailed explanation is presented in the results and discussion section As there was no animal husbandry within the study area, its pollution loads were not taken into account

River Flow Survey

Water of the Hadano basin, on its way from top of the alluvial fan to its center, infiltrates to the subsurface, flows as a ground water and springs out at the southern part

or gets stored deep under the ground (Ichikawa, 1978; Hadano City, 2003) Hence, to comprehend this kind of complex water circulation and its impact on water quality and quantity, the river flow survey was carried out in February 2010, on the basis of the

water temperature survey carried out in February 2009 (Shrestha et al., 2009)

The flow of each river was measured in detail from upstream to downstream at an interval of 200m to 300m River flow velocity was measured with a flow velocity meter (CM-1BN; Toho Dentan, Japan) at each monitoring point River flow was calculated by

Table 3 - Unit loads of pollution

Cultivated Land (kg/ha/day) 0.073 0.189 0.002

Urban Area (kg/ha/day) 0.293 0.044 0.005

Forest (kg/ha/day) 0.100 0.012 0.001

multiplying the river cross-sectional area by the flow velocity at various points along a transect across the main rivers Depending upon the circumstances and characteristics of rivers, river flow was measured at short distances also, such as at sections where the inflow of spring water and the ground water gushing out from river bed, banks and cliff walls were observed as well as at those points where river merges with its tributaries Accurate locations of monitoring points were digitally recorded by using Mobile GIS (ArcPad and GPS) (Fig 10) Measurements were done two to three times at the points where rivers merge with one another and at the respective rivers before their confluence

On the basis of these measurements, calibration was done Furthermore, measurements were done in clear weather conditions to avoid any unsteady conditions

Simulation for the Transport of Reactive Nitrogen in the Atmosphere

Reactive nitrogen compounds affecting water system and ecosystem that had been emitted to the atmosphere as a result of anthropogenic activities are mainly ammonia and nitrogen oxides These compounds, which are diffused from their respective sources, are transported along with wind and again get deposited to the earth surface either directly as dry deposition (gases and particles) or indirectly as wet deposition after getting mixed with rain or mist On the basis of these deposition mechanisms, the

simulation was undertaken using EAGrid2000-JAPAN database (Kannari et al., 2004; Kannari et al., 2007) to calculate the distribution of anthropogenic nitrogen deposition

in 1-km mesh for Kanagawa Prefecture

In this simulation, AIST-MM model (Kondo et al., 2001) developed by AIST (National

Institute of Advanced Industrial Science and Technology) was used to calculate the transport of reactive nitrogen compounds released from their respective sources to the atmosphere This model is based on the equations of meteorological models with hydrostatic and Boussinesq approximations, implying that the atmospheric layer is thinner as it extends horizontally and the range of temperature change is comparatively smaller Elevation data and land use information were also considered in the model to incorporate the effects of ground geography and land features, respectively Similarly, physical parameters such as heat capacity, reflection coefficient and thermal conduction

Muro Shijuhase

of soil layer were incorporated in the model These information are essential to calculate the sensible heat flux from the earth surface to the atmosphere that drives the atmospheric motion

Transportation of reactive nitrogen compounds was calculated with mass conservation

equation EAGrid2000-Japan (Kannari et al., 2007) was used for the emission inventory

data In this simulation, transformation of the compounds from gas to particle phases was simply taken into account with an exponential decay function of time The conversion of nitrogen oxide (CNOx) to its particle phase (CNO3-) with time is based on the following equation (Kitabayashi and Yokoyama, 1984; Environment Agency Air Quality Bureau, 1997)

CNO3- = CNOx • AN • {1-βexp(-KtN t)}PkNOx (v)

where AN: conversion coefficient from NOx to NO3-

β: initialization rate of NOx (= 1)

KtN: conversion rate from NOx to NO3-

PkNOx: remnant rate of particle material after subtracting the amount that

sublimes to gas phase

As already mentioned, reactive nitrogen gets deposited to the earth’s surface through dry and wet depositions For the calculation of dry deposition, the deposition velocity was introduced Usually, deposition velocity is a function of aerodynamic resistance, surface resistance and residual resistance Although the first depends on surface roughness and the last two depends on chemical species, constant deposition velocities were used for the sake of simplicity for all the compounds

vd = 0.002 (ms-1) day-time

vd = 0.0007 (ms-1) night-time where vd: deposition velocity

For the calculation of wet deposition, the amount of rainfall is necessary However, the differences of distribution and amount between those calculated by numerical model and observation are usually not small Then, Radar-AMeDAS Analysis rainfall data available in 1-km mesh was used

h: thickness of rainfall layer

Jo: rainfall amount in one hour (mmh-1)

(vi) (Environment Agency Air Quality Bureau, 1997)

Fig 11 - Results of atmospheric nitrogen simulation in GIS (month of July)

The simulation results for each month were first converted into GIS layer data in the mesh form and overlaid with the delineated drainage basins in ArcGIS (Fig 11) Then using the clip tool of ArcGIS, meshes with the values of atmospheric N deposition were extracted for each drainage basin Atmospheric N deposition load in each basin was calculated by adding the values of each mesh that lies within the respective basins, for the months when the water quality was monitored and the average value was taken in the analysis By the mapping of atmospheric N deposition loads in GIS, the distribution

of its loads was found to agree with the real situation trend-wise, such as higher deposition loads along roads, highways, ship passage and urban areas Calculated NOx were compared with observed values in Yokohama City The calculated concentrations were found lower than those observed However, the variation trend of their monthly average values was almost similar (Kondo, 2010)

RESULTS AND DISCUSSION

Survey Results

Water Quality Survey

It was found that the distribution of water quality in different time periods followed almost similar trends with some fluctuations at the upstream, midstream and downstream sections of rivers However, water quality concentrations observed in the month of January were found comparatively higher in most of the cases because of the unpredicted rainfall on the monitoring day in January (Fig 12) Particularly, the water quality of downstream monitoring stations located at tributaries and service water canals showed some higher fluctuations with higher concentrations mainly in January, such as stations 17 (Shimo-ochiai), 23 (Shimo-oduki) and 22 (Neshita-yohsui) in the case of TN and TP (Fig 13) COD concentrations at stations 17 and 23 were observed as high as 17 mg/L and 20 mg/L, respectively, in January (Fig 14) Similarly, it was found that the highest TN concentration was 19 mg/L at station 22 and the highest TP concentration

Monitoring Station Hadano Basin Dratnage Basin River Atms N (g/m2/month) 0.003 - 0.027 0.027 - 0.042 0.042 - 0.058 0.058 - 0.075 0.075 - 0.096 0.096 - 0.120 0.120 - 0.144 0.144 - 0.172 0.172 - 0.211 0.211 - 0.320

was 1 mg/L at station 23 in January In TDS, not many fluctuations were observed except at station 23 with a quite higher standard deviation

Hence, in all cases of COD, TN, TP and TDS, higher fluctuations were mainly observed

at station 23 Water quality was found distinctive as per the peculiarity of corresponding drainage basins However, TN concentrations were found consistently higher in drainage basins which were more urbanized than upstream basins (Figs 12, 13) This

Fig 13 - Water quality at each station including mean and standard deviation

(monitoring stations are shown in Fig 2) Fig 12 - Water quality variation of Muro River in different time periods