The 2006 NIMSNAMOS Lake Fulmor Deployments

1

• The 2006 NIMS/NAMOS Lake Fulmor Deployments
• The NIMS and NAMOS teams
• Center for Embedded Networked Sensing
• www.cens.ucla.edu

2
Outline

• Introduction NIMS and NAMOS
• History Data from prior Fulmor deployments
• Planning November 2005 March 2006
• Deployment 1 May 8-12, 2006
• Deployment 2 June 19-23, 2006
• New technologies and algorithms
• Plans for deployment 3 (August 06)

3
The NIMS System
4
NIMS RD Schematic
5
Hydrolab Sensor Node

• NitrateAmmoniumConductivitypHTemperatureFluo
  rescence (Chlorophyll A)
• Depth
• Attitude (pitch/roll/yaw)Compass Heading3 Axis
  Velocity

6
The NAMOS System

• Boat
• Fluorometer
• Temperature
• Multiple buoys
• Fluorometer
• Temperature at 6 depths

7
The Study Site

• Lake Fulmor is a small lake located in the San
  Jacinto Mountain range (altitude 5000 ft.) in
  Southern California with a maximum depth of 6m
  and low flow rates (1 cm/s). Relatively strong
  discrete wind events can have a great effect, and
  in summer prolonged wind events lead to formation
  of surface cyanobacterial scums.

8
(No Transcript)
9
Observations from 2005

• Seasonal changes Relative chlorophyll
  fluorescence increased an order of magnitude from
  May-October 2005
• Thermal stratification in Lake Fulmor changed
  throughout the year
• deepening of the surface mixed layer as the year
  progressed
• increasing the influx of nutrients into surface
  waters

10
Hypotheses from 2005

• Diel variations Maximum fluorescence measurements
  were observed at
• 1m depth from dusk -gt dawn (1800-0600) on a
  diel cycle
• Cycle was most pronounced during the July
  October deployments, during which cyanobacteria
  were highly abundant
• Likely that planktonic cyanobacteria are
  vertically migrating to optimize photosynthetic
  capabilities.

11
Plan Full Lake Assessment by Combining NIMS and
NAMOS

• Objective
• Map dynamic growth and migration of phytoplankton
• Direct measurement of biological phenomena in
  large scale 3D environment
• Solution
• Multi-scale methods
• Adaptive sampling
• Schedule
• March, May, July, August

12
Iterative Transect Placement Design (T 12-24
hrs)
13
Adaptive Sampling at the Surface (T 1-2 hrs)
14
Adaptive Sampling within a Transect (T 1-2 hrs)
15
Nature does its thing..
From David Caron Date Mon, March 27, 2006
There are now a couple of major considerations
here. I am worried About the rain/snow mix, and
using computers on our shore-based station in
that weather.
Date Mon, 20 Mar 2006 From Mike Taggart Okay,
here's the latest report from JR/Lake Fulmor as
of 1015AM on Monday 03/20/06 There is 6 to 8"
of snow around the footpath at Lake Fulmor.
While walkable, it is quite slippery and does NOT
look passable for a 2WD vehicle. There is a
thin sheet of ice covering about 70 of the lake
itself.
Date Mon, 27 Mar 2006 From Gaurav Sukhatme
Yes, the more I think about it, the more I lean
towards this strategy. Why not scrub this week
and wait for the next opportunity to go ?
Date Mon, 20 Mar 2006 From William J. Kaiser I
agree completely, we have to postpone. Thank you
Mike for this update.
Date Thu, 6 Apr 2006 From Gaurav Sukhatme
Bill - the weather does appear to have
worsened. I like your suggestion of simply
targeting May 8 as the first deployment.
Date Wed, 22 Mar 2006 From Gaurav Sukhatme I
hope the weather continues to cooperate
16
Deployment 1 May 8, 06 May 12, 06

• 7 buoys
• 1 boat
• NIMS RD transect
• 1 helicopter
• The NIMS and NAMOS teams

17
Chlorophyll Spike

• Chlorophyll spikes in early am by an order of
  magnitude (as seen by the NIMS shuttle)

18
Buoy Chlorophyll Timeseries

• Buoys on either side of the NIMS transect
• Daily chlorophyll variation is significant (peak
  in evening)
• Chlorophyll drop around 10 am on 5/9 -
  repositioning of the fluorometer on each buoy

19
Buoy Temperature Timeseries

• Buoys on either side of the NIMS transect
• Daily temperature variation is exaggerated at the
  top of the lake, non-existent at the bottom

20
Water Samples from May 2006

• Vertical water samples for chlorophyll, plankton
  community composition, dissolved particulate
  nutrients, and molecular microbial toxin
  analyses were collected at 3 buoy locations.
• In vivo chlorophyll concentration with depth at
  three buoy stations at 1400 (T1) 0100 (T2),
  respectively. The chlorophyll maximum appears to
  move deeper at night.

21
Chlorophyll at Transect
22
Temperature at Transect
23
The Effect of Smoothing

• Running median smoothing
• Cyan - unsmoothed

24
Deployment 2 Jun 18, 06 Jun 23, 06

• 7 buoys
• 2 boats
• NIMS RD transect
• First 24 hr run in old location
• Second 24 hr run in same location

25
Buoy 114 temperature NIMS scan overlay
26
Buoy 109 temperature NIMS scan overlay
27
Temperature Offsets between two Hydrolabs
28
Temperature is Slow to Settle
29
but PAR is not
30
PAR
31
Old vs. New Hydrolab
32
Old vs. New Hydrolab
33
Lake Mosaic to Estimate Biomass
34
Boat Localization using Stereo
35

• Design using local linear regression

36

• Multi-robot informative path planning with each
  robot starting from a different location, sensing
  temperature at Lake Fulmor.

37
NIMS-RD Overview of changes for August

• Adaptive Sampling (AS)
• Physical Sampling (PS)
• Sensor data via Subscription
• NAMOS/NIMS Aug 28th
• Basic Sampling Options that can be determined by
  data flow (first pass at AS)
• One-shot Physical Sampling that is user initiated
  (first pass at PS)
• Real-time feedback from sensors (done)

38
The Team

• Laura Balzano
• Maxim Batalin
• Henrik Borgstrom
• Peter Borgstrom
• Alan Butler
• David Caron
• Victor Chen
• Karthik Dantu
• Amit Dhariwal
• Andre Encarnacao
• Eric Graham
• Mark Hansen
• Brett Jordan

William Kaiser Jonathan Kelly Yeung Lam Vinay
Malekal Steffi Moorthi Koto Norose Carl
Oberg Henry Pai Srikanth Saripalli Beth
Stauffer Michael Stealey Gaurav Sukhatme Mike
Taggert Bin Zhang