TOTAL SUSPENDED SOLIDS LOADING OF STORMWATER FROM DIFFERENT LAND USE AREAS IN AN URBAN WATERSHED

Total suspended solids loading of stormwater from different land use areas in an urban watershed


Introduction.
Most cities in Sub-Saharan Africa, Nigeria inclusive, have witnessed rapid urbanization with the attendant complexities in land-use development in the last two decades [1].Changes in land-use pattern have led to significant alteration of the natural hydrological systems and the quality of water bodies including stormwater in urban areas [2].Urban transformation has altered the once forest dominated landscape into built-up structures and impervious surfaces like roads, walkways, bridges, airports, and parking lots.These hard surfaces radically decrease infiltration thereby exposing the landscape to greater volume of runoff [3].Anthropogenic activities additionally introduce nutrients, litter and biological matter to the catchment [4], which are easily swept away by stormwater into the water bodies at the downstream end of the drainage pathway [1].Suspended solids, rock particles, nutrients, trash, leaves, pesticides, trace metals, petroleum products, Escherichia coli, and faecal coliform bacteria are usually found in higher concentrations in stormwater and rivers within urban catchment than in natural systems [5; 6].These pollutants which are largely from diffuse sources pose serious problems such as growth of plants, depletion of oxygen, sedimentation, and change in colour to the quality of urban rivers.
Water quality parameters in various aquatic systems have been closely linked to the proportion and intensity of activities of the different land-uses within the watersheds [6].Land-use within a watershed impacts water quality by mediating the process by which diffuse pollutants such as suspended solids and nutrients enter the waterbodies [7].Managing stormwater is complicated by the extreme variability associated with runoff patterns from different land-use types under different rainfall conditions [6].In many parts of Nigeria, river water abstraction for household usage is an essential source relied upon by households and water vendors [5], so the protection of supplies against pollutants from diffuse sources is of great public health importance.But little attention has been paid to diffuse pollution and particularly stormwater quality in Nigeria and other parts of Sub-Saharan Africa.Most cities in southern Nigeria have high mean volumes of precipitation, and great number of rainy days leading to more complexity in the management of runoff [5].The mean transported load of solids is large because of poor environmental cleaning/solid-waste management and high volume of runoff [1].In most Sub-Saharan cities (Aba inclusive), the main task of the government, town planners and engineers is the channelization of stormwater to minimize flooding [1].The drainage infrastructure is often designed to convey stormwater out of the settlements and to downstream rivers and streams regardless of the global best management practices.Currently in Aba, greater emphasis and funding are given to flood mitigation with comparatively little attention paid to pollution risks of surface water.
The total pollutant load entering a river channel is a combination of suspended solids and pollutants dissolved in the water column, as well as pollutants associated with gross solids which may be floating litter and debris, or bed-loads rolling along the bottom of the conveyance system [8].Solids are picked up by stormwater runoff through erosion of pervious soils.Solids also enter the runoff stream from litter and trash on paved surfaces, household wastes, debris from construction sites, and through atmospheric deposition [7].Build-up and wash-off are key processes that act on urban surfaces to directly introduce suspended solids into stormwater runoff.Suspended solids are small particles that are floating in water as colloid, or suspended due to the motion of the water.Generally, the amount of particles that suspend in a sample of water is called Total Suspended Solid (TSS) [9].Total suspended solids load from urban surfaces is considered a major indicator of water quality [9].On the basis that TSS is a surrogate measure of urban stormwater quality [10], this study selected suspended solids as the main pollutant for consideration.
Increased concentrations of suspended solids in rivers and streams can have adverse impacts on the overall biological diversity of the water body [11].The greater the TSS in water, the higher its turbidity and the lower its transparency, making the water to appear murky [12].A high level of turbidity limits light penetration in water and affects aquatic plant growth as well as reducing the aesthetic appeal of waterways [12].Increasing load of suspended solids degrades water quality by increasing nutrient levels, organic materials, and metals, which are detrimental to the survival of aquatic organisms [7].Some studies have also identified heavy metals introduced by informal anthropogenic activities in urban spaces as a major contributor to the degradation of urban streams and rivers, with the most prevalent being iron, zinc, copper, lead, cadmium, and chromium [13; 14].The biggest threat to the quality and integrity of the rivers in most Sub-Saharan cities may be increased runoff pollution.Currently, there is dearth of empirical data on the TSS loading of stormwater or the concentration of suspended solids related to storm events in this region.
The TSS load of stormwater from a watershed and the size distribution of those solids depend on catchment land-use, the extent of construction activities in the catchment, and their interaction with the hydrological factors [15].Studies have shown that urban land-uses affect TSS loading of waterbodies in two primary ways: i) through intensive human activities which introduce wastes, particles, and organic matter to the landscape; and ii) through the build-up of paved and impervious surfaces leading to reduction in infiltration and increased volume of runoff during rainfall events [7; 8; 16-19].Understanding the relationships between land-use and water quality is very important for identifying major threats to water quality [16], [17].The understanding becomes very relevant in targeting intensive land-use areas and to institute necessary measures to mitigate pollution loading [18].But the level of pollution caused by urban stormwater runoff to the receiving waterbodies is largely unknown in low-income countries [20].Moreover, hydrological aspects vary geographically, so, extrapolating data on pollution load measured from other regions may introduce uncertainties.
The objectives of the study are: i) to assess pollution-load (for TSS and some trace metals) in runoff samples collected during storm events in six sub-catchment locations representing the six different land-use areas in the Aba-river watershed; and ii) to determine the effects of land-use on TSS loading of stormwater using Event Mean Concentration (EMC).The need for this study derives from the fact that most rivers in Sub-Saharan cities are heavily polluted, and Aba-river in south-eastern Nigeria is one of such [21; 1].The study will help to engender policy reengineering to tackle both surface water quality and flood risk concerns posed by rapidly growing urban land-uses, and to encourage proper watershed management.
Materials and methods.The Study Site.Aba, a city in south-eastern Nigeria was selected as a case study of cities in Sub-Saharan Africa where studies on diffuse pollution of surface water is still in the early stages.Aba is located within latitudes 5 o 04'N and 5 o 10' north, and longitude 7 o 20' E and 7 o 30' east.The city occupies an area of 72 km 2 (Figure 1) and lies within the humid tropical rainforest zone.The mean annual rainfall of Aba is high, with figures ranging between 2350 mm and 2850 mm [1].peaks in July and September.The most prominent landmark in the city is the Abariver, a tributary of the Imo-river, which traverses the city approximately in a diagonal section from the north-western end to the south-eastern border.Aba has a relatively low lying topography.The highest elevation is by the north-central part and lies 72 m above sea level while the lowest height is by the river valley which is approximately 36 m above sea level [22].The vast expanse of the city falls between 40 m and 50 m above sea level, and these areas are prone to flooding [22].The relief of Aba is such that the floodplain gently slopes from the east and from the west towards the Aba-river, which naturally drains the city (Figure 2a).The flood generated in the city is mostly channelled to the river, making it susceptible to much pollution.Aba has evolved organically over the years, with little town planning input.The resultant land-use pattern is complex, making inadequate room for open spaces, urban drainage systems, and preservation of floodplains for proper stormwater management.In the city, slums with very dense land occupancy practically eliminate the open spaces [1].Urban drainage in Aba is also adversely impacted by uncontrolled growth with excessive land development ratio, increased rates of soil imperviousness, and encroachment on floodplains and natural water courses [22].In recent years, the hydrologic environment has been under pressure from combined effects of physical development and heavy rainfall [22]; and the city has been heavily inundated, resulting to ubiquitous deluge of street surfaces [23].
Determination and Delineation of Land-use Sub-catchments.A GIS-based (Geographical Information System) watershed delineation tool was used to digitize the drainage area for each sampling point.Land-use/land-cover map layers and spatial analysis were used to determine the percentage imperviousness, land-use proportion and total land area in each sub-catchment.The sampling sites were selected by a careful consideration of the topography of the city to determine where the runoff tributaries suspected of high sediment contributions to the Aba-river converge (Figures 2a, 2b).
Sampling sites were decided based on representative stormwater flow path (Figure 3), accessibility, proximity in the watershed based on land-use, and safety.Some of the protocols followed in the characterization of the study site include: the use of GIS to delineate the Aba-river watershed into different land-use zones; and the identification of the drainage flow paths within each land-use zone and subcatchment.Following this protocol, six sampling stations (Figure 3) were established to monitor the stormwater quality.

Figure 3. Overlay analysis of wire frame (3D model) and flow vectors of Aba
Source: developed by the authors.
The stations are listed in Table 1.The study area exhibits mixed land-use pattern which was classified into six major zones: Residential Suburban (RSU) zone; Residential Urban (RUB); Industrial (IND); Residential Low-Commercial (RLC); Residential High-Commercial (RHC); and Residential Agricultural (RAG) zone (Figure 4).Stormwater Sampling.Stormwater quality characteristics were determined for the six sub-catchments in both low and high flow conditions.Stormwater samples were taken from the sampling stations established at selected runoff outfalls to the Aba-river in each of the six sub-catchments.Grab sampling was use in collecting the samples using PVC scoop, which received the flow and was then transferred to the sampling bottles.Multiple grab samples (composite sampling) and flow measurements are required over the duration of a rainfall event to determine Event Mean Concentration (EMC) values [24].A rainfall event is defined as any occurrence of rain, preceded by at least 10 hours without precipitation that results in an accumulation of 0.10 inches or more [25].Composite samples provide a better estimate of average concentrations.Flow-weighted composite samples were made using samples of equal volume taken at incremental intervals of flow volume and are combined.A sampling location free of debris, and a slow sampling rate were used to avoid capture of particles not flowing in the water.
Stormwater monitoring was conducted over a four-month period (June -September 2021) with three rainfall events monitored per month thereby capturing data for twelve rainfall events (12 samples) for each of the six land-use zones (n = 72) in the Aba-river watershed.A total of 72 flow-weighted composite samples were collected and analysed.Sampling was carried out simultaneously at the six sampling stations, by six different research teams.The research teams mobilized to

Journal of Innovations and Sustainability
ISSN 2367-8151 2023, Vol. 7, No. 2 https://is-journal.com the site as quickly as possible within reasonable time to capture the first flush at the onset of rainfall.Data collected for a rainfall event which did not significantly cover the entire study site were discarded.Sampling of a minimum of seven storms is adequate to determine annual TSS load trends to within acceptable limits of accuracy [26].Studies have shown that while conducting sampling to establish EMC values, the number of flow-proportioned aliquots to be taken over the full duration of the runoff event should be maximized as much as practicable [24; 27; 28].During each rainfall event, we collected between ten to twenty aliquots of grab samples at the interval of five minutes beginning from the first flush, and spacing through the entire duration of the storm.We prepared the composite samples which were then shipped to the National Root-crop Research Institute Umudike, Nigeria for water quality analyses.The laboratory provided a complete water quality analysis of each composite sample (throughout the research duration) that consists of the following water quality characteristics: TSS, and some trace metals (Zn, Cu, Pb, Cd, Fe, and Cr).The water quality analyses were generally conducted according to standard methods specified in the American Public Health Association (APHA) [29].
Data for rainfall events that produced significant runoff during the study period for the city of Aba were collected to determine EMC.Rainfall depth was collected directly with MET Office 5 inches Standard Rain Gauge (MET, United Kingdom).Runoff discharge was measured as the product of the wetted cross-sectional area of the drainage infrastructure (from where samples were collected) and the flow velocity.Flow velocity was measured using a flow meter -Model 801 EMflat (VALEPORT, St Peter's Quay, United Kingdom).Cross-section width of drainage channels was determined with measuring tape; and the wetted depth with a calibrated rod.Runoff from the catchments in the city is generally conveyed via rectangular drainage infrastructure, and canals.The following datasets were collected for each rain event; details can be seen in Appendixes 1-6.
i. Rainfall depth (mm) = average rainfall depth for the area measured for each event greater than 2.5 mm.
ii. Antecedent dry period (days) = time period since previous rain event greater than 10 hours between storm.
Laboratory Processing and Analysis of Samples.The filtration method [29] was used to analyse the samples for suspended solids.For the determination of TSS concentration, one-litre sample was extracted from each of the composite samples; agitated by hand to thoroughly mix the sample, and was filtered through a preweighed filter paper of pore size 1.5 micrometres [30].The filter papers were then dried at 105ºC for one to two hours and weighed with scientific scale, and readings were recorded to the nearest milligram.To determine the concentration of trace metals, the study employed the double beam (DW-AA320NR) Atomic Absorption Spectrophotometer (AAS) which used the flame test method.The concentration of TSS and trace metals were calculated using appropriate formulae as documented in American Public Health Association (APHA) [29].
Data Analysis.The Event Mean Concentration (EMC) and the Site Mean Concentration (SMC) were computed to analyse the water quality data.Event Mean Concentration is a widely used method of estimating stormwater pollutant loads.It is defined as the total constituent mass discharged during an event, divided by the runoff volume [31] as in Eq. ( 1): where M is total mass of pollutants (mg) over the entire event duration; V is total volume of flow (m 3 ) over the entire event duration, t is time (minutes); Qi (t) is the time variable flow (m 3 min −1 ); Ci is the time variable concentration (mg L −1 ); Δt is the discrete time interval (minutes) measured during the runoff event.
The SMC is the average EMCs of all storm Events.The SMC is taken to be the suitable measure of central tendency of the EMC's for a particular watershed or landuse sub-catchment [31].
Geometric mean was used to summarize TSS concentrations as the measure of central tendency due to the skewness of the data.The mean values of the measured parameters were compared with the standard set by World Health Organization [32] using one-sample t-test.Analysis of variance (one-way ANOVA) was used to determine the variation in pollutants concentration in stormwater across different land-use types and across different months of sampling.The descriptive statistics, ANOVA, and t-test were done using Statistical Product and Service Solutions (SPSS) for Windows, version 21.0 (SPSS Inc., Chicago, IL, USA) and Microsoft Excel 2016.
Results and discussion.Land-use and Watershed Characteristics.The Abariver watershed was delineated into six sub-catchments based on approximate landuse characteristics (Table 2).Okpu-Umuobo RSU and Eziama Aba-north RUB fall at the upstream end of the Aba-river watershed.The Glass-force industrial layout and the East Road RLC neighbourhood form the midstream, while the Ngwa-road RHC zone and the Owerri-Aba RAG zone fall at the downstream.The Eziama Aba-north area has the highest rate of imperviousness with 86 %, while the Owerri-Aba area is the least impervious with 40 %.The rate of imperviousness of the core urban areas (Eziama Aba-north, River layout, and Ngwa-road east) were higher compared with the outskirts of the city.These core areas correspond with the Residential Urban (RUB), Residential High-commercial (RHC), and Residential Low-commercial (RLC) land-use zones respectively.The mean volume of stormwater discharged decreased from the city centre toward the outskirts, with the RUB (10.5 m 3 min −1 ) having the highest, while the lowest mean discharged volume of 4.3 m 3 min −1 were from the RAG zone (Appendixes 1-6).The gradient of the area fluctuates between 2.5 % in areas further away from the Aba-river, to about 10 % near the river.Storm drainage channels were constructed at both sides of the roads following the natural gradient of the area, with their outfalls at the Aba-river.The elevation of the watershed ranges from 35 m above sea level along the river bank and towards the southern part, to 70 m further away from the river and towards the north.The Glassforce industrial area occupies the least land-use area of about 3.6 km 2 , while the Owerri-Aba neighbourhood has the largest land-use area of about 11.6 km 2 .Rainfall Parameters of Monitored Events.Twelve rainfall events were monitored between the months of July and September 2021 for each of the six sub-catchments in the Aba-river watershed (Table 3).The maximum rainfall depth in Aba within the period was 18.5 mm recorded on the 6 th day of sampling, while the minimum was 4.9 mm recorded on the 8 th day.Rainfall duration varied between 36 minutes (0.6 hr) to 7 hrs.The most intense rainfall recorded was 12.3 mm hr −1 while the least rainfall was 2.1 mm hr −1 .The antecedent dry period fluctuated between 12 hours to 8days.The details of the computed EMC for TSS and trace metals across all the landuse zones can be seen in Appendixes 1-6.The results show that SMC for TSS found in stormwater from the RHC land-use area was highest (1153.4mg L −1 ), followed by RLC land-use (834.2 mg L −1 ); IND land-use (588.3 mg L −1 ); RUB (525.7 mg L −1 ); RSU (481.3 mg L −1 ); and RAG land-use (216.5 mg L −1 ).For the trace metals, the mean Zn loading of stormwater slightly varied from 0.01 mg L −1 to 0.08 mg L −1 across the land-use areas, but with the industrial land-use recording the maximum EMC (0.09 mg L −1 ).Similar results were recorded for Cu, Pb, Cd, Cr and Fe.The mean Cu loading varied between 0.001 mg L −1 and 0.017 mg L −1 across the land-use zones, while the industrial land-use area turned out the maximum EMC (0.042 mg L −1 ).The mean Pb loading ranged between 0.001 mg L −1 and 0.006 mg L −1 .For Cd, the SMC recorded were less than 0.001 mg L −1 for all the land-use categories.The mean Cr loading were between 0.002 mg L −1 and 0.01 mg L −1 with the maximum EMC of 0.018 mg L −1 seen in the industrial land-use zone.The mean pollution load for Fe ranged between 1.4 mg L −1 and 5.6 mg L −1 across the land-use areas, with the maximum EMC of 8.4 mg L −1 recorded in the industrial land-use area.The pollution load for trace metals generally showed minimum variation across the land-use zones except for industrial land-use where comparatively higher EMC figures were recorded.Further comparison of the concentrations of TSS and trace metals in stormwater across different land-use areas were performed using One-way ANOVA (Table 5).The results indicate that there was a statistically significant difference in TSS loading between the different land-uses (F = 42.510,P < 0.001).A Tukey post hoc test (Appendix 7) revealed a statistically significant variation in TSS loading between RSU land-use and RHC land-use (P < 0.001), with a mean difference of (-6.228 mg L −1 ), which means that the residential high-commercial land-use area contributed 6.228 mg L −1 more of TSS than residential-suburban land-use area.Similarly, we observed a statistically significant variation in TSS loading between RSU land-use and RLC land-use (P < 0.001) with mean difference (-3.013 mg L −1 ).The TSS loading from RSU and RAG land-use areas also showed statistically significant variation (P < 0.001) with mean difference (306.667mg L −1 ), which means that residential-suburban land-use contributed 306.667 mg L −1 more TSS to stormwater than residential-agricultural land-use zone.There was no significant difference in TSS loading between RSU and RUB; and between RSU and IND land-use zones respectively.The results also showed a statistically significant difference in TSS loading between RUB land-use and RLC (P < 0.001) with RLC contributing 2.859 mg L −1 more of TSS; and between RUB and RHC land-uses (P < 0.001) with RHC contributing 6.074 mg L −1 more of TSS.Likewise, there was statistically significant difference in TSS loading between RUB and RAG land-use areas (P < 0.001), with RUB contributing 362.9 mg L −1 more of TSS.The Comparison of TSS loading from RUB with that from IND land-use area, and RLC with that from RHC land-use area showed no significant differences respectively.Alternatively, there was statistically significant difference in TSS loading between RLC and RAG land-use zones (P < 0.001) with RLC contributing 608.0 mg L −1 more of TSS; and between RHC and RAG land-use zones (P < 0.001) with RHC contributing 929.5 mg L −1 more of TSS to stormwater.Alternatively, the TSS loading of the residential-agricultural land-use area was significantly lower than the others.A comparative analysis of TSS loading for the four-months of sampling period (Appendix 8a) indicates a fair distribution of TSS load across the months following rainfall pattern, with means (612.2 mg L −1 ; 625.6 mg L −1 ; 640.5 mg L −1 ; 652.8 mg L −1 ) for June, July, August, and September respectively.One-way ANOVA was further applied to test the significance of the variation in TSS loading (Appendix 8b), and the result was not statistically significant (F = 0.372, P = 0.774).
The SMC for the trace metals were compared (Figure 6), and it indicates low concentration across the different land-use zones.However, we observed a spike in trace metals loading of stormwater from the industrial land-use zone.The heavy metal loads in stormwater for the land-use zones recorded significant differences: Zn (P < 0.001); Cu (P = 0.005); Pb (P = 0.006); Cd (P < 0.001); Cr (P < 0.001); and Fe (P < 0.001) (Table 5), but post hoc analyses revealed that the observed differences were primarily driven by pollutant loads from the industrial land-use zone.A Tukey post hoc test showed a statistically significant variation in Fe loading between the industrial land-use and each of the other five land-use zones respectively, but no significant difference was found between the pairwise comparison of the other five land-use zones.Similar results were observed in the Tukey post hoc test for Zinc, Cadmium, and Chromium respectively.Alternatively, there was no statistically significant difference in the loading of Copper across all the land-use zones (post hoc tests for the metals are available on request).
The comparison of the SMC for TSS and trace metals with the maximum permissible loading stipulated by the WHO [32] was carried out and the TSS loads were above the WHO limit for all the land-use zones (Figure 7).This was confirmed by a one-sample t-test which showed statistically significant results: RSU (t = 6.28,P < 0.001), RUB (t = 8.07, P < 0.001), IND (t = 14.64,P < 0.001), RLC (t = 19.97,P < 0.001), RHC (t = 21.44,P < 0.001), and RAG (t = 4.72, P < 0.001) (Appendix 9).The pollution load for all the trace metals across the six land-use zones were below the maximum permissible limits by the WHO.The of mean volume of stormwater discharged and TSS loading showed strong correlation indicating how runoff is an important conduit for the transport of TSS and the associated pollutants into the rivers and streams.
Discussion.This study found high EMC and SMC of TSS in stormwater across the six land-use zones but with significantly higher TSS loading from the residentialcommercial land-use zones; a low concentration of heavy metals across the land-use zones but with a spike for some heavy metals from the industrial zone; and mean runoff discharge volumes that follow the pattern of surface imperviousness in the study area.There was comparatively higher rate of imperviousness in the core areas of Aba (RHC, RLC, IND, and RUB land-use zones) because of the high rate of urbanization in the city [4] with the attendant compact physical development.The mean volumes of stormwater discharged in the study area followed the pattern of surface imperviousness in agreement with previous research findings which noted that hard surfaces radically decrease infiltration thereby exposing the landscape to greater volume of runoff [2].Some other scholars have also noted a correlation between rate of imperviousness and volume of stormwater discharged from urban catchments [7; 15].The natural gradient of the watershed also plays a role in mediating the rate of stormwater/pollutants discharge as observed in the industrial layout and Ogborhill neighbourhood east of the Aba-river.
The Event Mean Concentrations of TSS found in stormwater across the six landuse zones were higher than the maximum permissible loading approved by the WHO.These results are consistent with the findings of other scholars in this region such as [33] and [34], who also found high TSS concentrations in stormwater above the WHO permissible limits.This was the case because, there has scarcely been any deliberate effort to control stormwater pollution in Nigerian cities [20], and this is the primary justification for this study.Two sub-catchments in the Aba-river watershedthe commercial neighbourhoods (RHC and RLC) have been found to have significant contributions to the TSS load in the watershed.The EMC and SMC of TSS found in stormwater from the residential-commercial land-use areas were comparatively higher because, these areas have higher rates of imperviousness, more compact physical development, and greater intensity of human activities.This result agrees with the findings of E. Stein [6], and can be further interpreted to mean that rate of imperviousness, and in other words, intensity of physical development (in consideration with other factors) affect TSS loading of stormwater.The observed differences in TSS loading of the commercial land-use zones in the study area suggest that much of stormwater pollution and by extension deterioration of the rivers could be traced to the commercial districts of the city.These results corroborate the

Journal of Innovations and Sustainability
ISSN 2367-8151 2023, Vol. 7, No. 2 https://is-journal.comfindings of W. Dodds & M. Whiles [15] who observed significant correlation between urban commercial land-use and TSS loading of stormwater, and those of D. Jiao et al. [17] and S. Faiilagi [18] who also noted that urban land-uses affect TSS loading of water-bodies through intensive human activities which introduce wastes, particles, and organic matter to the landscape.Aba is largely a commercial city with lots of open markets.The volume of litter, dust, and rubbles generated in the commercial districts of most towns in this region is high due largely to the informal nature of activities.With poor waste management practices and poor environmental culture of the residents [35] litter is common sight resulting in high volume of gross pollutants and TSS in storm-water runoff.
There was a fairly uniform distribution of TSS in stormwater across the months of the sampling duration as the observed differences were not statistically significant.This means that the observed differences in TSS loading of stormwater in the study area may not be attributed to monthly variation in weather parameters.This was because, build-up and wash-off processes, and the effects of first flush were more important than climatic variability in determining TSS load in stormwater.This result corroborates the findings of D. Jiao et al. [17] who noted that the volume of TSS in stormwater is primarily driven by the build-up and wash-off of paved and impervious surfaces in urban landscapes.The mean volume of stormwater discharged and TSS loading were highly correlated indicating that runoff is an important conduit for the transport of TSS and the associated pollutants into the rivers and streams.This finding is explained by the fact that environmental sanitation and waste disposal practices are poorly managed in this region making surface runoff the primary vehicle by which dust particles, litter, oil and grease are washed off.In some cases, residents have had to throw trash directly into street runoff during rainfall.This is in agreement with the findings of C. Ogbonna et al. [1] who observed that the mean transported load of solids in Nigerian cities is large due to poor environmental sanitation and high volume of runoff following rainfall events.
The TSS loading of the Residential-Agricultural (RAG) land-use zone was found to be comparatively low.This result however conflicts with the findings of W. Dodds & M. Whiles [15] who found a positive correlation between agricultural land-use and TSS loading in stormwater.These scholars noted that farming activities aid soil erosion which increase sediment yield in storm runoff.However, the variation in our study may be attributed to lower density of physical development and availability of vast expanse of secondary forests and open spaces in the RAG zone of our study site, which aid infiltration and lower the volume of runoff during rainstorm.This study observed a relatively flat topography in the RAG zone (which reduces the intensity of soil erosion), and less volume of runoff discharge compared with the other land-use zones.In addition, littering is less rampant in the RAG zone as the nearby farmlands play a significant role in the traditional methods of waste disposal.These results suggest that with lower intensity of physical development and more green areas or forested open spaces, the quantity of stormwater runoff will be
The observed trend of low SMC of heavy metals across the different land-use zones (except for the industrial land-use zone which recorded a spike for some trace metals) was due to the fact that the major source of heavy metals in surface water in this region is industrial sewage, with minor contributions from rock weathering, automobile transportation, and mechanic workshops as noted by M. Adeyemi et al. [36].The fact that most of the manufacturing industries in the study area are located in the IND zone explains the spike in trace metals loading from that zone.Though the EMC levels for the trace metals fall below the maximum permissible standard by the WHO, there is need for closer attention to be paid to the effluent from industries in Aba as the results show that some of the industries may have been channelling their untreated effluent to the Aba-river via the storm drainage system.
Conclusions.This study examined TSS loading of stormwater from different land-use areas in a river watershed using the city of Aba in Nigeria as case study.The authors observed high concentration of TSS in stormwater across the six land-use areas, in excess of maximum permissible limits by the WHO.The study also found statistically significant variations in TSS loading of the residential-commercial landuse zones, and the residential-agricultural zone compared with the rest of the land-use areas in the city.While the commercial land-use zones yielded higher TSS loading compared with the other land-uses, the residential-agricultural zone yielded lowest TSS, suggesting that commercial land-use areas of towns in this region contribute significantly to TSS yield in stormwater, and accentuates surface water pollution.Beyond TSS, the concentrations of heavy metals in stormwater were comparatively low but with statistically significant difference across land-use zones.However, the observed differences were mainly driven by the IND land-use as there was consistently higher concentration of trace metals in stormwater discharged from the industrial layout.This calls for further study of the management and discharge of industrial effluents in the city.
The quality of stormwater in the Aba-river watershed is poor by reason of high TSS loading; hence the rivers are under heavy pollution.The residents are therefore, advised to desists from abstracting water from the rivers for domestic uses unless it is subjected to standard purification.As a way of mitigation, the city authorities should use the instrumentality of town planning regulations to recover open spaces that have been invaded by informal activities especially in the commercial districts.This will help to improve infiltration, reduce the volume of storm runoff, and improve sanitation and waste disposal.The city authorities in this region should invest more in waste management and urban drainage, and create the enabling environment for the private sector to develop and manage standardized infrastructure for stormwater treatment and flood control.These would reduce environmental pollution, improved stormwater quality and by extension the quality of the rivers.

Figure 1 .
Figure 1.Location map of Aba city, in Abia State, Nigeria Source: developed by the authors.The rainy season begins by March and lasts till early November, with bimodal

2a 2b Figure 2 .
(a) Contour map of Aba main town; (b) Digital Terrene Model of Aba Source: developed by the authors.

Figure 4 .
Figure 4. Land-use map of Aba Source: developed by the authors.
Figure 5 graphically summarizes mean TSS loading of stormwater from the different land-use zones and reveals that commercial land-use areas (RHC & RLC) contributed more TSS to stormwater than the other land-use areas in the watershed.

Figure 5 .
Figure 5. Mean TSS loading for land-usesSource: developed by the authors.

Figure 6 .
Figure 6.Trace metals loading of different land-use zones Source: developed by the authors.

Figure 7 .
Figure 7. TSS loading of land-use zones compared with World Health Organization's limitsSource: developed by the authors.

Table 2 Land-use characteristics of the Aba-river watershed
Source: calculated by the authors.

Table 3 Summary on rainfall data of events monitored
Physical Parameters of Stormwater Quality Across Land-use Zones.The Site Mean Concentrations (SMC) were computed from the respective EMC values for TSS and trace metals for each land-use area, and are summarized in Table4.
Note. *figure is average for the Aba-river watershed.Source: calculated by the authors.