GEOG 566

         Advanced spatial statistics and GIScience

June 12, 2017

Nitrate Concentrations in Space and Time and how they relate to Red Alder Cover in different watersheds

Filed under: 2017,Final Project @ 10:38 am

My Spatial Problem Blog Post (Revised)

By: Nicole Feiten

Research Description

I am exploring aquatic chemistry among 10 streams in the Oregon Coast range. I will be running anion analysis over the period of 1 year sampling monthly. These streams are divided into three categories, three are located East of Mary’s Peak in the Rock Creek Watershed, three near the top of Mary’s peak (highest in elevation), and four streams are sampled on the Oregon Coast within 200 meters of the Pacific Ocean. I would like to perform a comparison analysis of anions, especially nitrate, between the site locations and find if there is any significant difference in anion abundance between the site locations and over time (6 months of data).

Research Questions

Is there a relationship between nitrate concentration data and site locations?

Do the differences in nitrate concentrations indicate a pattern between site locations, and do they show a pattern over time?

Is there relationship of nitrate data and Red Alder tree cover?

Is there a relationship between nitrate concentrations and land use practices?
Dataset Description

The dataset I will be analyzing includes temporal anion data (for 6 months, project still in progress) and GPS points for all 10 site locations. Additionally, I used NLCD land use data sets and Red Alder Cover data sets from the Oregon Spatial Database Library. My spatial problem focused on nitrate concentration data collected at the 10 sites and the relationship between sites spatially and temporally.

The temporal scale for this analysis includes one sample per month for a six-month period.

Spatially the sites are located some distance apart, listed in the table below in Kilometers. The distances below reflect Euclidean Distance, stream distances were not used, as not all sites were not connected systematically.



Distances Rock Griffith Middle Fork Kaiser Yew Crooked Cape Gwynn L.Cumm Cumm
Rock 1.6 1.91 6.44 8.76 11.62 58.07 58.78 59.04 58.7
Griffith 1.12 5.23 7.63 10.65 56.62 57.41 57.67 57.27
Middle F. 6.17 8.61 11.7 57.23 58.09 58.23 57.9
Kaiser 2.47 5.61 51.85 52.54 52.79 52.36
Yew 3.2 49.89 50.47 50.71 50.24
Crooked 47.65 48.32 48.35 48.09
Cape 1.29 1.66 1.47
Gwynn 0.399 0.422
L. Cumm 0.338

Table 1 Distance between sites in Kilometers



I hypothesize that nitrate concentrations will be similar in sites that are closer together and more different the further apart they are located. This is a classic scenario of autocorrelation. I also suspect that stream nitrate concentrations will be higher in lower elevation streams than in higher elevations streams. Looking further, upon my Red Alder percent coverage analysis, I suspect sites with a higher percent cover of alder will have higher concentrations of nitrate in the streams.  Lastly, I will analyze land use in each of the watersheds, I suspect that watersheds that have higher amounts of agriculture and other anthropogenic activity will have a higher concentration of nitrate.

Analytical Approaches

Initially I will perform the analysis in Microsoft Excel, I am interested in seeing if sites that are closer together show similar concentrations of nitrate. First I calculated Euclidean distances in ARC GIS using the measure tool. After I obtained those distances I will use the “Correl” function in Excel to correlate known nitrate concentrations at each site (an average for the 6 months) and the distances apart the sites are.


Table 2 Correlation of nitrate concentrations and site distances

Figure 1 Correlation Analysis showing clustering of sites that are closer together


Next I will perform another analysis in excel I will sort the concentration data between streams within varying locations. This will allow me to find the concentration of the nitrate data between points (stream sites) and analyze their relationships over time.

Figure 4 Concentration of Nitrate over time in the Coastal stream sites

Finally, I will use the Oregon Spatial Database to clip layers of relating to percent alder cover in the watershed and landuse practices to identify a relationship between the nitrate concentrations and alder cover and land use.


Table 1 Nitrate Concentrations and Percent Alder Cover

Watershed Area (m2) Alder (m2) Percent Alder Cover Average Nitrate
Cape Creek 4353700 3537.74 0.081% 1.88
Crooked Creek 31760400 3226.39 0.010% 1.35
Cummins Creek 21388800 222.89 0.001% 1.44
Gwynn Creek 2865000 5371.57 0.187% 1.61
Little Cummins Creek 3056600 361.35 0.012% 1.72
Yew Creek 9786600 969.92 0.010% 1.01
Kiser Creek 100500 1017.97 1.013% 0.23
Rock Creek 38375500 2359.26 0.006% 0.19
Middle Fork Creek 3294000 339.23 0.010% 0.17
Griffith Creek 4434800 2021.86 0.046% 0.13

Figure 5 Relationship between percent alder cover and nitrate concentration



In addition to the above figures I have produced distance and correlation tables, as well as scatter plots of data to analyze for relationships. Below are examples of maps I produced in ARC by clipping land use cover and watershed size for each stream sites. Alder concentrations were produced by joining the attribute tables for the alder layer and the watershed area shape files. Even though there wasn’t as strong of a correlation of alder cover and nitrate concentrations, further research may show different results if just stream riparian area is used.

I found higher differences (indicating higher nitrate concentrations) in the mainstem rivers than the tributaries. Additionally, I saw higher concentrations of nitrate in the spring samples as compared to the winter samples. I expected higher values for nitrate in the mainstem, such as rock creek due to the volume of water coming from upstream. Naturally the smaller tributaries would exhibit lower concentrations than the larger order stream that catches the two. I also expected nitrate differences and concentrations to decrease in the spring time. If the inputs of nitrogen to the system remain relatively stable, then the uptake length of nitrate (the most sought after in stream nutrient cycling) would shorten and cause concentrations to decrease.

Overall nitrate concentrations over the past year in the ten watersheds are relatively low (~0.1-1ppm), and the results obtained were not particularly surprising.  There was no clear pattern linking red alder cover and nitrate concentrations in these watersheds (Table1). Additionally, there were no surprise land use features in the watersheds, and thus no real link to nitrate concentrations in this analysis. Most of the land use in these watersheds are occupied by mixed and evergreen forests, neither are particularly linked to elevated nitrate concentrations. I would say this analysis was consistent with my expectations, starting with low nitrate concentrations, and suspecting most of the land use would be forest cover I predicted there likely not to be a strong link between land use and nitrate concentration. However, I expected slightly more red alder cover in the coastal watersheds and was surprised how little that cover related to the concentration of nitrate.


Figure 6 Coastal Watersheds Landuse

Figure 7 Coast Range Watersheds Landuse


Scientific Significance

Baseline data collection for anions in Oregon Coast Range streams over a year has not been documented before. In the face of a changing climate, it is important to document and characterize stream chemistry for Forest managers in the future. Increases in nitrates in these streams can affect local flora, fauna as well as increases nutrient loading to the ocean. This loading can affect macroinvertebrate and salmon habitat which can alter ecosystem dynamics and services that currently exist.


What I learned about Software and Statistics

I have basic knowledge of R, but I think it would the most useful tool for statistical analysis of the data. Arc-info is a program I at one time used daily but it has been many years since. I would like to be able to map the anion data between the 10 sites but may struggle in the beginning linking the data with Arc. I have no experience with Python or other relevant spatial analysis software.


This class taught me more about spatial autocorrelation, especially because my data were grouped in this way. Since I performed the majority of my analysis in Excel, I didn’t expand my knowledge in ARC GIS as much as I had hoped. However, I did learn that if you don’t need sophisticated software to produce your analysis, it is OK to use what you already know. As I move forward in my research I am looking forward to studying multivariate methods and using the software PC-Ord.



Comment Feedback

Tutorial 1 Comments

The most useful comments from Tutorial were about having a more structured step by step process in my design. I agree that if this tutorial were to help someone in the future complete the same, or similar analysis, that more detailed instructions would be useful. I updated my tutorial to include step by step instructions in the methodology of my analysis.


Tutorial 2 Comments

In tutorial two I was able to clearly relay the knowledge and importance of my study. Like most feedback I was pleased to hear about additional analysis regarding clusters of data points that may point me in a direction to understand more about Alder cover and nitrate fixation in streams. Perhaps a buffer layer around the sites and a clip of the cover to just the buffer would show me a better picture about the linkages between red alder and nitrate concentrations. In the future, I would like to continue to use the alder layers, with a buffer added and potentially perform analysis that involve other concentrations of Anions as well.


Comments from Dr. Jones


The comments I received from Dr. Jones were overall positive and insightful. Exercise one I received feedback on the potential for nitrate to be transported via dispersal or diffusion which I was unable to justify given the data set that I was working with. Also, I found a negative correlation between Rock creek and it’s neighboring creek. I decided that Rock creek with a higher volume of water and larger watershed area, would have likely produced a different concentration of nitrate compared with the other two.


Exercise two comments included thoughts about the equation I chose, which was difficult for me to understand as I have not used equations to describe data in the past. The distances between the sites were measured using a Euclidean calculator in ARC. As well as an explanation of higher concentrations in Rock Creek due to Higher concentration of water coming from upstream, I realize this now is incorrect and what I actually meant is it is contributed to by the larger volume of water in Rock creek and additionally the larger watershed.



The third exercise brought a question about how alder cover may not be associated with nitrate concentrations. After thought, I realized that maybe it is not the percentage of Alder cover in the entire watershed that may affect stream nitrate concentrations, but potentially just the trees in the riparian buffer zone. In future work, I would like to clip buffer zones around the stream and analyze both alder cover and land use cover in just this area and run a correlation with stream nitrate concentrations.





Print Friendly, PDF & Email

No Comments

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

© 2019 GEOG 566   Powered by WordPress MU    Hosted by