A CAREER of STATISTICS in AGRICULTURE, ECOLOGY and the ENVIRONMENT - PowerPoint PPT Presentation

1 / 65
About This Presentation
Title:

A CAREER of STATISTICS in AGRICULTURE, ECOLOGY and the ENVIRONMENT

Description:

A CAREER of STATISTICS in AGRICULTURE, ECOLOGY and the ENVIRONMENT – PowerPoint PPT presentation

Number of Views:80
Avg rating:3.0/5.0
Slides: 66
Provided by: nscottu
Category:

less

Transcript and Presenter's Notes

Title: A CAREER of STATISTICS in AGRICULTURE, ECOLOGY and the ENVIRONMENT


1
A CAREER of STATISTICSinAGRICULTURE, ECOLOGY
and the ENVIRONMENT
  • N. Scott Urquhart
  • Senior Research Scientist
  • Department of Statistics
  • Colorado State University

2
ILLUSTRATIONS FROM A STATISTICIANS WORK
  • Many ways a statistician can contribute
  • Work as a team member
  • Help in many ways to get good data
  • Assist in analysis and interpretation
  • Personally rewarding
  • Contributed to solving relevant societal
    problems
  • Interesting, even fun
  • On-site work to understand problems
  • Investigators really appreciate site visits
  • Financially adequate

3
WHAT IS NSU INVOLVED IN RIGHT NOW?
  • Im mainly an administrator now
  • But still do some data analysis and research
  • People implementing environmental monitoring
    programs
  • Want to know how good a study will be
  • Even though data will be gathered for years
  • Domain From multiple watersheds to the Western
    States
  • Measures
  • Water quality From acid rain responses to
    bug population
  • Forest health
  • Bird populations in the National Parks
  • Summaries
  • Status how big is the response
  • mainly as a cumulative distribution function
  • Trends How is the response changing thought
    time?

4
EVALUATING PERFORMANCE OF A MONITORING DESIGN -
HOW?
  • Power Probability(detecting a trend of a
    specific size)
  • Prob(Rejecting HO for a specific point in HA)
  • Variance of estimated regression coefficients
    can be decomposed in the contributions of
  • River drainages,
  • Specific sites (randomly selected)
  • Year effects
  • Unexplained remainder
  • Illustration bull trout population sizes in
    Montana
  • Collaborator Monitoring scientist at the
    National Marine Fisheries Service, La Jolla, CA
  • NSU statistical models worked illustration
  • Colleague software for doing power
    calculations in this
  • situation

5
POWER TO DETECT TRENDVARYING YEAR EFFECT AND
TEMPORAL DESIGN
6
MAJOR PROJECTS NSU HAS BEEN INVOLVED WITH
  • Sugar beet investigations
  • Rural development
  • Beef cattle feeding
  • Commodity futures hedging
  • Breeding onions for pink root resistance
  • Using waste products Sewage sludge Cesium
    138 gt ?
  • Catching
  • Deer in a net
  • Desert bighorn sheep
  • Deer fawns
  • Electric meter testing
  • Using computers to assist in learning

7
MAJOR PROJECTS NSU HAS BEEN INVOLVED WITH
continued
  • Sampling sales tax records
  • Program evaluation for the Grand Canyon
    Monitoring and Research Center (GCMRC)
  • Designing and implementing a vegetation study
    for GCMRC
  • Designing studies for EPAs Environmental
    Monitoring and Assessment Program (EMAP)
  • Arctic contaminants Research Program
  • Designing vegetation studies in Arctic Alaska

8
MAJOR PROJECTS NSU HAS BEEN INVOLVED WITH
continued
  • Sampling sales tax records
  • Program evaluation for the Grand Canyon
    Monitoring and Research Center (GCMRC)
  • Designing and implementing a vegetation study
    for GCMRC
  • Designing studies for EPAs Environmental
    Monitoring and Assessment Program (EMAP)
  • Arctic contaminants Research Program
  • Designing vegetation studies in Arctic Alaska

9
PROGRAM EVALUATION FOR THE GRAND CANYON
MONITORING AND RESEARCH CENTER (GCMRC)
  • Objective
  • Review the near-river terrestrial research
    program of GCMRC
  • External panel
  • Biologists quantitative scientists
  • Academic and federal employees
  • From Glen Canyon Dam through the Grand Canyon,
    nearly to Lake Meade
  • Evaluation panel of 8
  • 11-day raft trip through the Grand Canyon
  • Accompanied by 9 people who had conducted
    research along the river

10
MAP OF THE GRAND CANYON AREA
11
THE PANEL SUPPORT CREW
12
THE BEGINNING
13
THE REAL START _at_ LEES FERRY
14
CAMPED LIKE ANY OTHER TRIP
15
WE SAW MAJESTIC SCENERY
16
WENT THROUGH RAPIDS !!!!!!!!
17
THE END WAS IN SIGHT
18
_at_ THE END
19
A CLOSER LOOK _at_ VASEYS PARADISE
20
THE END OF A HIKE FROM TAPEATS
CREEKTHROUGHSURPRISE VALLEYTODEER CREEK
21
WE SAW DESERT BIGHORN 4 PLACES - ANOTHER POSSIBLE
IMPACTOR
22
DESIGNING AND IMPLEMENTING A VEGETATION STUDY FOR
GCMRC
  • Objective Locate and lay out vegetation
    transects
  • NSU selected 100 potential sites
  • Randomly, subject to spatial restrictions
  • Each potential site was
  • Determined to be vertically faced no transect
  • Or the transect was laid out and documented
  • Result
  • 20 sites to be visited annually
  • 40 sites to be visited only once every three
    years

23
NO VEGETATION HERE!!!(MILE 135.2)
24
ROCKY, BUT VEGETATED SITE _at_ 12.3
25
A SANDY SITE _at_ 171.5
26
A VERY ROCKY SITE _at_ 95.0
27
THE CREW AFTER TWO WEEKS ON THE RIVER
28
DESIGNING STUDIES FOR EPAs ENVIRONMENTAL
MONITORING AND ASSESSMENT PROGRAM (EMAP)
  • Distinctive EMAP perspective
  • Define the population of interest
  • Conduct a probability survey of IT
  • Carefully define the sampling frame
  • Variability probability selection of sites, but
    with spatial balance
  • Carefully define the responses to be evaluated
  • Train field crews well
  • Manage the data well while maintaining a good
    audit trail
  • Learn from past mistakes, throughout
  • NSU has been very involved with the process for
    evaluating bugs in streams

29
CADDIS FLIES
30
MESA CREEK (ORC 20)
31
SURBER SAMPLER IN USE
32
COLLECTING IN A STREAM
33
FIELD CLEANING OF A COLLECTION
34
BAGGING PRESERVING A COLLECTION
35
BAGGED COLLECTION - LAB
36
EVALUATING THE EVALUATION ERROR
  • Here, there is no way to repeat the process of
    pulling clumps apart gt Repeated measures
    are impossible
  • Alternative Split collections
  • 5050
  • Separately evaluate each half
  • Examine similarity of splits to study
    evaluation error

37
COLLECTIONEMPTIED INTO A BEAKER PRIOR TO
SPLITTING
38
PLACING A COLLECTION IN THE SPLITTER
39
SPLITTING A COLLECTION
40
COMPLETED SPLIT
41
SPLIT COLLECTION READY FOR ENUMERATION
42
MAP OF ALASKA
43
EPAs ARCTIC CONTAMINANTS RESEARCH PROGRAM
  • Reality Arctic haze has increased greatly over
    the past 30 years.
  • Apparent origin
  • Northern Europe
  • Including the former Soviet Union.
  • Arctic haze contains lots of bad stuff
  • Question
  • Has much of this bad stuff been deposited in
    Arctic Alaska?
  • Approach
  • Use lichens as a natural accumulator
  • Collect specific lichens across a possible
    gradient of sites.
  • NSU Selected general site location developed
    local site selection criteria

44
BEGINNINGS OF TRIP INTO THE ALASKA NATIONAL
WILDLIKE REFUGE BETTLES, AK
  • WHERE IS THIS FLOAT PLANE SUPPOSE TO TAKE OFF???

45
APPROACHING SCHRADER LAKE IN ANWR
46
FLOAT PLANE LEAVING US IN AT SCHRADER LAKE IN ANWR
47
EVALUATING VEGETATION IN ANWR
48
THE WAY OUT FROM SCHRADER LAKE
49
DESIGNING VEGETATION STUDIES IN ARCTIC ALASKA
50
TEAMS LEARNING ABOUT PERMAFROST
51
VIEW FROM OUTSIDE - REFRIGERATED ENTRANCE
52
IN THE TUNNEL
53
AN OLD ICE WEDGE
54
A SEGMENT OF THE TRANS-ALASKA PIPELINE
55
A HILLSIDE ON THE NORTH SLOPE
OR IS THAT ALL?
56
MUSKOX ALONG THE HAUL ROAD
57
RESEARCH SITE AFTER JUNE SNOW
58
LOOKING TOWARD THE RESEARCH SITE FROM INSIDE THE
BROOKS RANGE
59
TYPICAL VALLEY BOTTOM (NOTE MOOSE IN CENTER)
60
GENERAL GOALS OF R4D
  • Develop ecophysiology models of plant growth on
    the North Slope
  • To support more general models to allow
    evaluation of likely impact of energy-related
    development
  • NSU On-site experimental and sampling design
  • For many studies

61
(No Transcript)
62
HILLTOP AND NEARBY STONE STRIPE
63
STONE STRIPES AND EQUIPMENT FOR MEASURING
PHOTOSYNTHESIS RATE
64
POWER FOR MEASURING PHOTOSYNTHESIS RATE
65
EQUIPMENT FOR MEASURING PHOTOSYNTHESIS RATE
Write a Comment
User Comments (0)
About PowerShow.com