Kawa Stream TMDL Project

Water quality characteristics of Kawa Stream

REPORT INTRO WQ Standards Suspended solids Conductivity Nutrients

The discussion presented on this page pertains to Kawa Stream above the estuary (located entirely within Bay View Golf Course). A description of Kawa Stream includes photographs and definitions of the various reaches. Also, a series of photographs taken along the stream (including a map) are provided by Oceanit. Links to photographs of the water quality monitoring stations are listed on the AECOS project page.

Water Quality Standards

Hawaii's water quality standards are contained in Chapter 54 of Title 11 of the Hawaii Administrative Rules. The standards are also reproduced on each of the parameter discussion pages linked in the discussion below. In our discussion, measurements made regularly as part of a monitoring program with minimal bias to conditions influencing stream flow (i.e., the HDOH 1999-2000 monitoring program) are used to assess compliance with the standards. Measurements made on an irregular basis (AECOS storm water monitoring) with particular emphasis on rainy periods cannot be used for compliance assessment purposes, but contribute to our understanding of where and how pollutants reach Kawa Stream.

Suspended Solids

Streams accomplish work in the watershed: they erode the land and move the loosened material downslope and, eventually, into the sea. This solid material is carried by the stream as either bed load, suspended particles, or floating debris. Measurement of suspended particles is done as either turbidity or TSS (total suspended solids) or both.

At stations on Kawa Stream where turbidity and TSS were monitored regularly over time (see 1992 and 1999-2000) both tended to increase with distance downstream, peaking (in these data) around Station 007 in the middle reach (Table 1). Below Station 007, the values tended to decrease as shown by values from Station 005 at the upper end of Bay View Golf Course. Particulates might be expected to increase as one moves down gradient in the watershed due to increased competence of the stream; that is, the increase in volume and velocity of stream water as a result of increasing drainage area, resulting in the stream's ability to maintain greater amounts of particulate matter in suspension. The marked decrease in both turbidity and TSS at Station 005 in the lower reach of Kawa Stream likely resulted from deeper, slow moving pool areas in the downstream waters where stream velocity is retarded and particulates settle out. Note that TSS between Station 007 and Station 005 decreased by about 50 percent, whereas turbidity levels decreased by only about 30 percent. These facts indicate that larger and heavier particles are settling out more readily than the very smallest particles that make the water cloudy (i.e., turbid).

Table 1. Turbidity and total suspended solids, geometric means
for the Kawa Stream 1999-2000 baseline (HDOH, 2001)
Turbidity TSS Turbidity/
TSS ratio
Sta. No.
(ntu) (mg/l)
Station 10 (= 008R) 1.8 1.8 1.0
Station 8 (= 008L) 5.7 3.9 1.5
Station 007 10.1 5.6 1.7
Station 6 (= 016) 8.9 5.6 1.5
Station 005 6.2 2.8 2.1

Grand means 6.8 4.1 1.7
std. dev. 2.9 - 16 1.6 - 10
n (data set) = 98 106 98

NOTE: Station 10 means based upon only 7 events (June-September 2000).

Geometric mean turbidity levels for the stream as a whole exceed both wet and dry season State water quality criteria (Table 2), while TSS concentrations are below the State's criteria. These differences raise the question as to whether or not the State criteria for turbidity and TSS are in synchrony; i.e., shouldn't both parameters either be in compliance with State criteria or not? The answer is "not necessarily".

Table 2. Wet and dry season turbidity and TSS geometric means for Kawa Stream
compared with State water quality criteria.
Turbidity TSS Turbidity/
TSS ratio
(ntu) (mg/l)
wet season samples
Kawa Stream meana
7.3 4.4 1.9
State criteria 5.0 20
dry season samples
Kawa Stream meana
6.4 3.9 1.5
State criteria 2.0 10

        a 13 dry season events and 12 wet season events, HDOH Stations 5, 6,7, 8, and 10

Of course, during high runoff periods, turbidity and TSS rise dramatically in Kawa Stream. Collections of samples from different parts of the stream, from the numerous drainage outlets, and from selected points over the time progression between the onset of rainfall through peak discharge and on until water quality returns to normal will be used to assess the effect of storms of various magnitudes on suspended sediment transport. Relating these kinds of measurements to water quality standards requires consideration of storm frequencies for the watershed. In the main channel, values easily exceed all of the turbidity criteria during significant rain storms. But the standards do not allow one or several measurements purposely collected during a storm to establish noncompliance.

Measurements and observations suggested that a significant source of turbidity and TSS in the stream during storms in 1999-2000 wet season came from a construction project underway at Hawaiian Memorial Park (see photo). This project straddled upper (intermittent) Kawa Stream (central branch) and included a newly constructed culvert. During the November through January 2000 period at least, steep, graded fill banks suffered significant erosion during monitored storms. Improperly designed silt fences were partly to blame. As shown in the May 2000 photo, the situation was eventually corrected. Hopefully, adverse water quality impacts decreased or disappeared once the slopes became vegetated. What is not known is how long soil deposited in the stream bed below the project by 1999-2000 wet season storms will take to flush completely out of the system.

Preliminary results from monitoring several storms suggest that turbidity (and perhaps TSS) in the stream are excessive following a heavy rainfall, but urban storm drains in this watershed may not be as significant a source as newly graded land.


Conductivity is a measure of the salt content of water. Typically, conductivity will increase in the downstream direction as flowing water picks up minute quantities of salts from the rocks and soils. Once the water reaches the estuary, conductivity rises very sharply as fresh water mixes with the high salt content (3.5%) sea water. Within the context of the Kawa Stream monitoring, this mixing point occurs between Stations 005 and 004.

Conductivity can decrease during periods of high flow caused by storms, because rainwater has a very low dissolved salt content. Any soluble salts on the ground at the beginning of a large storm are quickly picked-up in the surface flow, and become diluted by the rain. Our storm water sampling shows stream conductivity declines as stream flow increases due to rainfall run-off.


Although a number of chemical substances can be regarded as "nutrients," being essential for growth in living organisms, two classes of chemicals are of greatest concern from a water quality standpoint: compounds of nitrogen and phosphorus.

We consider first the data obtained from the HDOH stream monitoring program as summarized in Table 3 (see also HDOH Data Summary for additional statistics). These measurements, representing average conditions in Kawa Stream, reveal that high levels of nitrogen compounds are carried by the Stream. It is also evident that nitrate + nitrite (hereafter called "nitrate," but understood to include a small but unknown proportion of nitrite or NO2) is the dominant nitrogen form. Considering that nitrate salts are very soluble, the results suggest sources of these salts in the watershed regularly contribute nitrates to the stream. Their fate, in terms of delivery into Kane`ohe Bay, is discussed in the section on Kawa Stream estuary.

Table 3. Nutrient geometric means for the Kawa Stream 1999-2000 baseline (HDOH, 2001)
NO3 + NO2 NH3 Total N Total P
Sta. No.
(ug/l) (ug/l) (ug/l) (ug/l)
Station 10 1097 2 1204 11
Station 8 (= 008L) 812 10 1015 72
Station 007 613 51 908 27
Station 6 (= 016) 138 46 533 35
Station 005 997 24 1369 37

Grand means 538 23 928 37
std. dev. 175 - 1655 6 - 89 492 - 1747 19 - 71
n (data set) = 107 107 106 107

NOTE: Station 10 means based upon only 7 events (June-September 2000).

Nitrate concentrations are especially high in the upper reaches of the stream and tend to decrease somewhat with distance downstream through Station 007 in the upper middle reach. The lowest mean value (geomean = 138 ug/l) occurs in the middle east branch (Pohai Nani Branch) of Kawa Stream at Station 016. This branch drains subwatershed (Basin) 4. Because both flow and concentration are less than the main branch, this smaller stream contributes much less nitrate loading than subwatersheds (Basins) 1, 2, & 3.

Nitrate levels appear to rise again in the lower reach of the stream below Kaneohe Bay Drive (Station 005). As will be shown later, much of this increase in the lower reach appears to be contributed from the flow of the lower west branch (Castle High Branch) of Kawa Stream that arises in subwatershed (Basin) 5 draining Kaneohe Bay Shopping Center and parts of Castle High School campus. Concern that Castle High School might be contributing excessive nutrients led to a class project by students [see Castle High School, 2001].

In as much as the State's water quality standards have different criteria1 for samples collected in the wet season (months of November through April) as opposed to those collected in the dry season (months of May through October), calculating statistics from the HDOH data based upon "season" may be instructive. Table 4 presents a comparison of wet and dry season geometric means for Kawa Stream with State of Hawaii water quality criteria. The nitrate + nitrite levels exceed State criteria by 7x in the wet season and nearly 20x in the dry season (Table 3).

Table 4. Wet and dry season geometric means for Kawa Stream compared with State water quality criteria.
NO3 + NO2 NH3 Total N Total P
(ug/l) (ug/l) (ug/l) (ug/l)
wet season samples
Kawa Stream meana
486 27 939 41
State criteria 70 --- 250 50
dry season samples
Kawa Stream meana
586 20 918 33
State criteria 30 --- 180 30

a 13 dry season events and 12 wet season events, HDOH Stations 5, 6,7, 8, and 10

Ammonia concentrations are relatively low in the upper reaches of Kawa Stream and increase to a geometric mean maximum of 51 ug/l at Station 007 near the Namoku Street bridge before decreasing further on downstream. Total nitrogen distribution in the stream is similar to that for nitrate, since it is largely determined by nitrate levels in the stream.

At every station on Kawa Stream where water samples have been collected as part of a regular monitoring program, total nitrogen (TN) and nitrate + nitrite numbers exceed the respective water quality criteria1 for streams (mean values in red in 1992 and 1999-2000 data summaries). A comparison of wet and dry season geometric means for Kawa Stream with State of Hawaii water quality criteria (Table 4) reveals that total nitrogen means are well in excess of State criteria, but high total N is largely the result of high nitrate + nitrite levels in this stream. Considering individually each of the samples analyzed (and rejecting instances of nitrate analytically exceeding total N), nitrate is seen to constitute anywhere from 4 to 96% of the total nitrogen, but averaging 63% in the HDOH data set.

Historic data suggests that, on average, inorganic phosphorus (ortho-P) changes little over time in the lower reach of Kawa Stream. Total phosphorus,on the other hand, appears slightly elevated in the wet season as compared with the dry season considering both historic data and HDOH averaged data (Table 4). This follows from the fact that TP is clearly related to particulates carried by the stream -- particulates introduced by run-off. Figure 1 shows the correlation between TP and TSS for run-off influenced samples2. HDOH Station 8 (=08L), monitoring out-flow from the upper west branch of Kawa Stream, is characterized by mean TP levels that are double the levels measured at other stream stations. This branch is located within a forested gulch, although may arise within Hawaiian Memorial Park.

Figure 1. Regression of total suspended solids on run-off influenced TP concentration.

The situation with respect to nutrients in the estuary of Kawa Stream is now well established because of numerous sampling events conducted in that body of water since 1991 (see AECOS, MRC, DOH). As Kawa Stream water enters the estuary, it creates a situation where noncompliance with (geometric mean not-to-exceed) estuarine criteria for ammonia, nitrate + nitrite, total N, and total P (all of the nutrient standards) is the norm. HDOH data (1999-2000) show a trend of steadily decreasing TN in the seaward direction within the estuary. Inorganic nitrogen also appears to decrease along this gradient. These data are evaluated elsewhere for the influence of dilution with Kane`ohe Bay water within the estuary. However, the water exiting Kawa Stream (Station 001) is noncompliant (assessed against the standards for embayments) for nitrate + nitrite and ammonia only. While the Kawa Estuary is not much more than a linear, man-made channel, an environmental benefit of estuaries and coastal wetlands is buffering stream water, providing natural protection to marine resources.


1 -- The Hawaii water quality standards for streams include include separate wet season and dry season values or criteria for the nutrient parameters (nitrate + nitrate, total N, and total P), as well as having three "not-to-exceed" values based upon frequency of occurrence (a total of six standards for each of these parameters). The discussion here presently considers only the "geometric mean not-to-exceed" criteria.
2 -- "Run-off influenced samples" are explained elsewhere. No correlation was found between non-run-off influenced samples and Total P.