Jeffery S. Horsburgh

fallsnowpeasInternet and Web Development

Nov 12, 2013 (3 years and 9 months ago)

119 views

Sensors, Cyberinfrastructure, and
Water Quality in the Little Bear
River: Adventures in Continuous
Monitoring

Jeffery S. Horsburgh


Amber Spackman Jones, David K. Stevens

David G. Tarboton, Nancy O. Mesner


Three Breakout Topics


Designing continuous monitoring networks



Sensor network telemetry and communication



Integrating optical measurements with other
water quality data to improve predictions

Observing Infrastructure

Horsburgh, J. S., A. Spackman Jones, D. G. Tarboton, D. K. Stevens, and N. O.
Mesner (2010
), A sensor network for high frequency
estimation of water quality
constituent fluxes
using surrogates,
Environmental
Modelling

& Software
, 25,
1031
-
1044,
doi:10.1016/j.envsoft.2009.10.012
.

Designing Continuous
Monitoring Networks

“The Space Challenge”


How do water quality conditions vary throughout a
watershed?


As a result of hydrologic features?


As a result of different land use?


As a result of management practices?



What processes (human and natural) drive the
variability?


Sources

-

What are the sources of pollution and how much is coming
from each source?


Transport pathways
-

How do pollutants reach the water bodies in the
watershed?


Fate

-

what happens to the pollutants once they get into a water
body
?

“The Time Challenge”


How and why does WQ change over time

(minutes
-

years)


In response to natural events (seasons, storms,
snowmelt, etc.)


In response to human events (reservoir
management, diversions, return flows, etc
.)


Are WQ conditions getting better or worse?


What might happen in the future?


Climate change?


Land use change?


Little Bear River Sensor Network


7 water quality and
streamflow monitoring
sites


Temperature


Dissolved Oxygen


pH


Specific Conductance


Turbidity


Water level/discharge



4

weather stations


Air Temperature


Relative Humidity


Solar radiation


Precipitation


Barometric Pressure


Wind speed and direction


Soil moisture and
temperature at 5 depths



Spread spectrum radio
telemetry network





Water Quality Issues


Nutrients (Primarily P)


Sediment

Urban
Stormwater

Runoff

Agriculture

Wastewater Treatment

Pollution Sources

Objectives


Use
high frequency measurements of
discharge
and turbidity to better
quantify suspended
sediment and total phosphorus
fluxes



Design the observing infrastructure required to
enable high frequency estimates of constituent
fluxes using surrogates



Study how
high
-
frequency sensor data collected
at multiple sites
improve our understanding of
hydrology and water quality

Sensor Deployment


How do we deploy the
sensors so they are:


Representative


Secure


Lots of great guidance
out there


Every site is different!


Can constrain site
selection and network
design


Have you seen my
turbidity sensor?

It used to be
right here!

Location, Location, Location


Access?


Can you get permission
from the landowner?


Can you get there all
year long?


Does it freeze?


Cross section?


What sort of telemetry
options will work?


Power?

The Human Element


Huh… Why does the river all of the sudden
get deeper during the middle of the summer?


Site selection in network design


Your research questions
matter


the space and
time challenges


Sometimes the “right” site for the science isn’t
accessible


Detailed scoping is required, and every site is
different

Sensor Network Telemetry and
Communication

Why Telemetry?


The remote technician


I don’t have to go to
the field to check the status of my sensors!


Adaptive sampling


its raining at my weather
station and the stage has increased in the
stream, do I change the frequency of my
observations?



What can we do with data in real time that we
can’t do with offline data?



Telemetry Network Design


Which technologies
to choose?


Satellite


UHF/VHF/spread
spectrum radios


Ethernet


Land line telephone


Cellular telephone


Mixed networks


Considerations


E
quipment
cost


Regular service
cost


Service availability


Terrain


Vegetation


Distance


Required
bandwidth


Availability


Reliability


Power


Interference


Required
expertise



Radio telemetry
network setup


Optimal
placement of
radio repeaters
given monitoring
site locations

Viewshed

Analysis

Paradise

Repeater

Mountain Crest High School

Remote Base Station

Upper South

Fork Site

Lower South

Fork Site

Lower East

Fork Site

East Fork

Weather Site

Confluence

Site

UWRL Base

Station Computer

Key

Internet Link

Radio Link

Stream Monitoring Site

Climate Monitoring Site

Paradise

Site

0.8

2.9

0.6

2.9

1.3

1.9

5.2

Telemetry


Viewsheds

and radios have nothing to do with
hydrology and water quality



…but, if you want to network sensors or
have real time access to data you have to get
this expertise…

Data Integration

Observing Infrastructure

Horsburgh, J. S., A. Spackman Jones, D. G. Tarboton, D. K. Stevens, and N. O.
Mesner (2010
), A sensor network for high frequency
estimation of water quality
constituent fluxes
using surrogates,
Environmental
Modelling

& Software
, 25,
1031
-
1044,
doi:10.1016/j.envsoft.2009.10.012
.

Hydrologic Information Science

Hydrologic conditions

(Fluxes, flows, concentrations)

Hydrologic environment

(Dynamic earth)

Physical laws and principles

(Mass, momentum, energy, chemistry)

It is as important to represent
hydrologic environments

precisely with

data as it is to represent
hydrologic processes

with equations

Hydrologic Information Science

(Observations,
data models
, visualization

Hydrologic Process Science

(Equations,
simulation models
, prediction)

Slide from David Maidment

The Data Deluge

One day = 48 observations

One week = 336 observations

One month = 1440 observations

One year = 17,520 observations

Two years = 35,040 observations

Three + years = 50,000 + observations

Times 7 Sites = 350,000 observations

Times 10 + Variables per site = 3,500,000 observations

Plus different versions of the data (raw versus checked) = 7,000,000 observations

Plus 4 weather stations with 10 + variables = almost 12,000,000 observations

You need some infrastructure to manage and share the data.

http://hydroserver.codeplex.com



A platform for publishing space
-
time hydrologic
datasets that is:


Autonomous with local control of data


Part of a distributed system that makes data
universally available


Basis for Experimental Watershed or Observatory
data management and publication system


Standards based approach to data publication


Accepted and emerging standards for data storage
and transfer (OGC,
WaterML
)


Built

on established software


MS SQL Server,
ArcGIS

server


Open Source Community Code Repository


Sustainability



Ongoing Data Collection

Data
presentation,
visualization,

and analysis through Internet
enabled applications

Internet Applications

Point Observations Data

Historical Data Files

GIS Data

HydroServer

ODM Database

GetSites

GetSiteInfo

GetVariableInfo

GetValues

WaterOneFlow

Web Service

WaterML

Observations Data Model (ODM)

Soil

moisture

data

Streamflow

Flux
tower data

Groundwater

levels

Water Quality

Precipitation

& Climate


A
relational database

at the single observation level


Metadata for
unambiguous interpretation


Traceable heritage from
raw

measurements to
usable

information


Promote
syntactic

and
semantic

consistency


Cross dimension

retrieval and analysis

Horsburgh, J. S., D. G. Tarboton, D. R. Maidment, and I. Zaslavsky (2008), A
relational model
for environmental and water resources data,
Water Resources Research
,
44, W05406
, doi:10.1029/2007WR006392.

Data Values


indexed by “What
-
where
-
when”

Space, S

Time, T

Variables, V

s

t

V
i

v
i
(s,t)

“Where”

“What”

“When”

A data value

ODM


Supports:


different types of data and different needs


a number of different queries


you can slice and
dice the data however you want


Many analysis packages (MATLAB and R) can
connect directly to a database to get data


Supports data publication using the CUAHSI
Hydrologic Information System (HIS)

Loading data into ODM


Interactive ODM Data Loader


Loads data from spreadsheets and
comma separated tables in simple
format


Streaming Data Loader (SDL)


Loads data from datalogger files on
a prescribed schedule


Interactive configuration


SQL Server Integration Services
(SSIS)


Microsoft application accompanying
SQL Server useful for programming
complex loading or data
management functions

ODM Data Loader

SDL

SSIS

Managing Data Within
ODM
-

ODM Tools


Query and export



export data series
and metadata


Visualize



plot and
summarize data
series


Edit



delete,
modify, adjust,
interpolate, average,
etc.


Data Management and Publication
Cyberinfrastructure

Horsburgh, J. S., and D. G. Tarboton (2010), Components of an integrated
environmental observatory
information system,
Computers &
Geosciences, doi:10.1016/j.cageo.2010.07.003.

Horsburgh, J. S., D. G. Tarboton, M. Piasecki, D. R. Maidment, I. Zaslavsky, D.
Valentine, and
T. Whitenack (2008), An
integrated system for publishing
environmental observations
data, Environmental
Modelling

& Software, 24, 879
-
888,

doi:10.1016/j.envsoft.2009.01.002.

Wait a second


I’m not a computer scientist
!

Yes…but…


We are collecting more data


higher spatial and
temporal resolutions


The way we store and manage
data
can either
enhance or inhibit our analyses


Visualization and analysis of large datasets can be
difficult and require specialized software


You will need to share data



Are we training our students to work in a data
intensive environment?

Data Management Requirements


What are the 20 queries that you want to do?


e.g., “Give me simultaneous observations of turbidity
and TSS collected during the spring snowmelt period
so I can develop a regression in R.”



How will you organize and manage your data to
satisfy those queries?



What are the standards we will use as a
community to share data and metadata?

H
ow do Natural
F
eatures and Human

Activities
A
ffect WQ Conditions?

Spatial distribution of total suspended solids fluxes in the Little Bear
River
for
2008. The areas of the node markers are proportional to the total

suspended
solids
fluxes, which
are expressed in metric tons.

Questions?

Support:

EAR 0622374

CBET 0610075