Sita Karki
The objective of this study is to estimate the geothermal heat flux using the satellite imagery. During the process of heat flux estimation using thermal imagery, it is difficult to correctly quantify the heat flux due to the interference from the background environment. Thus we are calculating geothermal heat flux from geothermal features in Yellowstone National Park after quantifying the background heat due to variables such as slope, aspect, vegetation, soil etc. We are using Landsat Thematic Mapper for the study. We have defined non- geothermal areas outside Yellowstone National Park as model training areas. Considering the background variables, we have developed the multivariate regression model for the training areas. The multivariate regression model developed for the training areas will be used for estimating geothermal heat flux inside Yellowstone National park.
Research objectives:
Quantifying geothermal background data via multivariate analysis for anomaly identification in satellite imagery
Model training in "cold" areas near Idaho/Montana/Wyoming outside of Yellowstone
Model testing in Yellowstone
____________________________________________________________________________________________________________________
Activities for 4/21:
finish updates on final 4 graphs
start collecting input data for model run of ESRP to show lack of structured anomalous pixels in quiescent zone
continue with expanded Tendaho run
{SKN & CB: work on modifying text, expanding Coso section}
Activities for 2/24:
Make slides
Activities for 2/17:
contact CB about cutoff graticule, Tendaho geology
Revising Ch (1,2,4,5)
Prepare questions list
Activities for 2/11:
{SKN: finish 1,2,5}, Fig 18 histogram: put unit label on x-axis, remove orange maps,Fig 14: make 1 sdv bold, 2 stdev regular
Explain more about <273K data removal, including how often those points happened; How many data points were left total in the end?
Context image for Tendaho,Continue to harass CB for rift layer,Update Ch4
Activities for 2/3:
{SKN finish Ch 4 review, review Ch 1,2,5},prep histogram version of 2700m vs all-elevation figure (Fig 18)
Fig 14: remove extra lines on Opalescent, Nuphar, and Gray Lakes; change key to "Residual plus 303K", add 1 stdev whiskers (use one in bold, other in regular line),Discuss filtering used to remove outliers from USGS time series,Tendaho image: remove faults, remove wells, email CB daily until she sends rift layer,email committee about tentative date to get confirmation
Activities for 1/27:
Make figure showing that full model approach is basically the same as <2700m approach (use subtractions at various sites), add text explaining consistency of model (this is a good thing that it doesn't change a whole lot)
USGS temp fig: remove residual points, narrow range of y-axis, and 1stdev or 2 stdev bars to field data
add to Conclusions chapter. Include key point summaries for each of the earlier chapters, restate your problem statement, flesh out full paragraph on future directions, discuss changes from Landsat 5 to Landsat 7 on resolution and signal,Tendaho: remove wells from figure, add black dashed lines for rift traces, overlay on aerial imagery, Longer-term Tendaho: expand area to NW -> include more of Red Sea Rift, will include Carrie's InSAR to compare volcanoes,Revisit Intro Chapter for thesis, identify areas that need to be updated,{SKN reads/comments on draft}
Activities for 1/20:
finish results/discussion draft (in the same manuscript document as the intro content), make figure for updated Tendaho, Closeup view for <2700m YNP, USGS temp figure, Make bullet-point list of key things to include in Conclusions chapter, {SKN: poke Carrie about Tendaho}
Activities for 12/30:
Get the grades from Shannon, Register for classes, get computer access and vpn, Read two more Tendaho paper and Coso, Start writing Results and Discussion,Run the model with pixels elevation less than 2700m,Analyze USGS temperature data
Activities for 12/23:
Fig 3: start x-axis at 250,Fig 3&4: adjust y-axes so that they have the same max-min and same major label difference (consistent w/i fig, not between 3&4),Fig 6: use 100 m axis hash marks for both x-axis,Fig 7: add "pixels" to y-axis label; modify Residual x-axis to label at -40, -20, 0, 20, 40 (or without the 20 and -20, if too crowded); adjust x-axis for Raw to be structurally similar to that of Residual, Known Anomalies: remove horizontal guide lines, add "pixels" to y-axis label,Fig 8: zoom in on NW anomaly polygon & run again. Also: map showing extent of each of the zones used in the calculations A-E, be sure to include anomaly polygons, Can we get fault data (vector format) for Coso area? If so, overlay with current map,Search for papers on Coso, Tendaho,Read 2 Tendaho papers SKN sent,Mono Map: bracket shifted in correlation of map units, volcanic islands group,start updating Intro/Methods
Activities for 12/9:
Figs 3 & $ - double axes;Fig 5. Keep crosses but revert to equal interval for symbology; add upper bound to high temp and lower bound to low temp
Fig 6: label raw vs residual panels, remove pale gray background;Fig 6: what is the highest elevation of the known anomalies? Can you make a histogram of the elevations of the known anomalies?;Fig 7: increase color contrast; play with solutions to strange x-axis labels (perhaps divide into 2 parts);Fig 8: add panel "D" for a model run that only uses the number of pixels inside of the displayed area;Fig 9/10: We need to look up more about the Coso Geothermal field for our discussion;Fig 10: Digitize and filter points to minimize visual "noise" in the figure; display as colors using stdev-coloration (not 2 stdev filter);Both: read Kebede for Tendaho info;{remind Carrie};Mono map comments: - "Some unit exposures on the printed map..." -> "Units are labeled where possible, and unlabeled units are attributed in the database."; - "Site ID" -> "Geochemistry sample location; see text";- Shoreline in Explanation of map symbols: we need to get rid of lines for older shorelines and rename entry in map symbols as "Modern shoreline"; - Trachydacite of Northern Paoha flow -> label as 1 and 2 on map (there is already a dividing line between them on the map; 1 is the northern one); - "Clover ranch high stand" -> "Clover Ranch High Stand" also, add "(1962 m elev)"; - for Historic high stand, add "(1958 m elev)"
- need more color difference between Qnrf and Qnre; - Fig 8 caption: "amont" -> "among"; - Fig 1 caption: "edifces" -> "edifices"
Activities 12/2:
See if you can hunt down contact for time series data that only displays recent month on USGS webpage;Fig 3: Adjust y-axes so that they share the same vertical exaggeration (same axis length, same difference between Ymax and Ymin);Fig 4: add area of pool vs. residual temp (can put on same plot, but with separate axes for the different temperature ranges);Fig 5: try with equal distribution of pools instead of equal range (for symbology), convert triangles to crosses;Fig 6: x-axes need to have the same ranges (Xmax-Xmin);Fig 6: what is the highest elevation of the known anomalies? Can you make a histogram of the elevations of the known anomalies?;Fig 7: Use 0-850,000 for both Y-axes. Make bin size 10 for both plots; x-labels based on actual position instead of bin beginning and end; for residual plot, make sure label on pt 0;Fig 8: add panel "D" for a model run that only uses the number of pixels inside of the displayed area;Fig 9/10: We need to look up more about the Coso Geothermal field for our discussion
Fig 10: Digitize and filter points to minimize visual "noise" in the figure; display as colors using stdev-coloration (not 2 stdev filter);Both: read Kebede for Tendaho info;{Send Afar output to Carrie for comment};{Send the handful of (very) minor edits to Mono Map};{read/comment on Intro/methods}
Activities for 11/25:
{read/comment on Intro/methods};email YNP group about additional temperature time series data for Steamboat and/or other sites. Can we get coverage of dates that we did RS analyses for? (include GRC report; mention SKN and ISU, as well as thesis);BeloMerg Fig - remove x-axis label from upper panel, remove citation from lower x-axis label (citation will go in fig caption);RawEmittance Figs: Does raw emittance data actually occur in those segregated stripes?;histogram figs: not actually histograms, x-axis meaning is unclear. Re-check meaning of x-axis. You mentioned that the real histograms were odd looking -- I'd really like to see one, even in draft form;figures for Fig 7 (impact of area)
prep draft figures of CA and Ethiopia sites;For results/discussion sections: make a bulleted list of key points for the figures
Activities for 11/12:
R^2 figure: use 2 decimals right of zero;Break methodology flow chart into 2 separate pieces: training vs. testing
UpperUnclipped: label Old Faithful, correct "gyser", giant (?); increase font size of feature labels; are additional hot areas outside of bounds real?
Similar updates (hot zones, text size) to MidwayUnclipped;AveragedResidualEETMEmittance - remove uncolored entry from legend; cut one "River Group" label; cut a couple of group labels from figure to create space; Averaged...(zoom) - add 1 &2 to Violet Mud Pots; remove uncolored level from legend;Link BELO poolvsFieldTemp with BELO pool vs residual; turn 2 into superscript;Raw temp vs field temp: add units to axes, is correlation significant?; RawTempVsFieldTemp: check on best-fit line and r^2; add another panel showing short-term temp changes in at least 1 YNP geothermal feature (see park website for data access) [we want to tell a story about how variable data really is, and how much our method improves what we've got];ResidualvsBelo: add dashed line across graph at Residual = 0; check best-fit line and r^2 -- better than for raw?;ScatterPlot: make corresponding scatterplot for raw emittance vs. elevation;make histogram of residuals, histogram of raw
Activities for 11/4:
shasta is back and usable - draft Fig 7;update Intro/Methods;results thoughts: Mammoth, CA; Afar; COTM New Zealand or Iceland?
discussion: snow as a natural filter, snow as a problem for clear skies, how is this models different in useful ways from past models
Activities for 10/28:
{SKN: read/comment on new Intro/methods};make histogram of residuals for whole-park analysis (the one that looks like Watson's)
make a draft of Figure 7 (wait for shasta to come back);Fig 4: try plotting as different symbols all on same axes
Figs 1a & 1b: replace pixel outline symbology with filled symbology;Fig 5: add similar graph showing area of pool vs. field temperature
Fig 3: make x and y axes have same min/max/gridstep; Fig 2: add more focus locations, confirm Landsat 5 vs 7
using no more than 1p, show a flow chart for all the steps of the model; update Intro/Methods images per discussion (b/w, add redundant variables to r^2, etc.); using no more than 1p, show a flow chart for all the steps of the model;B/W R^2 - white text box approach
look for modified-NDWI to include in R^2 table;training/testing area map: make N in arrow bold and make training area designation letters bold
convert close-up images pixel colors to blue, yellow, red; allow pixels outside of Hutchinson boundaries to display; shrink inset map
Activities for 10/21:
incorporate edits for Intro and Methods, includes adding figures,Results/Discussion section: put together the figures
draft or better versions of as many as are already started, plus sketch versions on new figures; at least 1 sentence per figure about why it's important/or what it says, write a bulleted list of important points to cover in results/discussion
Activities for 9/23:
poster: shrink Fig 3, add aerial photo to Fig 2, plus explanatory labels of images, cut Fig 4,label Fig 5 with "Raw" "Background" and "Filtered" tags from Fig 4,enlarge Figs 6 & 7 if possible, format, check for spaces in author list, switch order, make department part of title block much larger;{finish Intro edits},incorporate edits once received
Activities for 9/16:
poster:reduce Intro text,cut figs 2 and 5;add images showing different variables (hill shade, NDVI, NDWI, elevation, slope, etc.) for 1 training site; give each image a title that explains purpose of image; add image of simple pixel subtraction; if space, add YNP photos (be sure to cite them); need authors, affiliations; need ISU Department of Geosciences logo along top (can get from website); add white boxes behind Figs 4 and 7 to get rid of green;% pixel image: put y-axis crossing x-axis in x's 0; adjust key so that high, medium, low (or the other way around); write key entries using format: High (>338K); Medium (310-338K); Low (<310K); {edit Intro},incorporate Intro edits;continue work on Methods section
Activities for 9/10:
Redo PCC_Pixel symbology using BeLo temps to differentiate between high/med/low. Include chosen temp ranges in figure legend.
{finish edits and return};work on edits;fill out problem statement worksheet;Write a thesis statement: what is the point/purpose of this paper?
make draft of GRC poster
Activities for 9/2:
update PCC_Pixel to use symbology to distinguish between low, medium, and hot pools; check Landsat8 use - compare 8 vs 5 for specific sites w/ closest possible days;if Landsat 8 good, run for melting road zone; {Read methods; comment};write additional methodology text that covers background from other manuscript
Activities for 8/26:
make chart showing residual vs. % of pixel that is pool;add another date to road melt suite;calculate residual for road melt images;continue on methods draft
Activities for 8/19:
look up pool sizes from BELO, see if relationship between size and residual;make table showing BELO temp and raw (not residual) temp at same position;storyboard of intended figures - make a complete figure list; prep clean draft versions of as many as possible;clean view of melting road zone in pre-2014 images? make tiled images showing residual temps there over time;continue on methods draft;write introduction draft
Activities for 8/12:
For BELO data table: convert to Kelvin, check extreme difference values for signal mixing in pixels;check WGS84 vs NAD27 for BELO points
check positive vs negative residual - which should be showing up as "hot" in final output (residual emittance seems v. negative)
histogram of (BELO-residualTemp) (in Kelvin);make up plots of BELO comparison data specific to feature type;continue writing - rough draft of methods; move Mono figures per sketch
Activities for 7/22:
Compile temperature data from BeLo11 into table, display in Arc; produce table with feature name, UTM, measured temp, our temp, area, edge?
clean up MIdway and Upper Geyser figures (remove yellow, make scale bars same size, "geyser," graticule along edges (no lines through), remove most site names and increase font size on the remaining ones);if time: pull recent Landsat, try to find the area with the melting pavement in YNP
Activities for 7/15:
Midway Geyser Basin: identify the named pools/geysers and measure their areas. Is there a size threshold for what we can see in our output?
Which can you see if you reduce filter to 2stdev? 1stdev? etc. Make list of features visible at each threshold, their area, and name (if available), along with coordinates; Make a figure with panels showing residual at 1, 2 and 3 stdev;If you finish Midway Geyser Basin with time, take on Upper Geyser Basin;If time: make figure showing outline of our anomalies and published blue anomaly outlines on aerial photography to illustrate what we are seeing in our output (hot springs vs. surrounding deposits)
Activities for 7/8:
{SKN work on MPI for MC code};Expand Landsat 5 vs Landsat7 search to include dates within 1 week spread; do subtraction on final images; if not enough, make composites;Collect good imagery dates for composites (will happen, regardless);Convert scales to be the same for ALL images this will mean that some images only use part of the lowest color block, due to different low-temps from standard deviation subtraction;start looking through JW's shared data, so we know what else to look for
Activities for 6/30:
{SKN work on MPI for MC code};contact JW for coordinates list, sizes and depths of features as appropriate;find dates with Landsat 5 and Landsat 7 coverage for direct comparison (2 dates?; if none, build composite outputs using 5+ dates for each sensor (composite as average value at each pixel);replot Landsat 7 tiled image to put on same scale as Landsat 7, add new color box for additional temperatures beyond 5's scale;<gap due to end-o-semester, travels>
Activities for 5/1:
for main image, restrict to Watson domain, shift legend onto Yellowstone Lake in image, darken shaded relief for better contrast. Use point values from Vaughn to compare with hot zones, but don't keep "point" features visible -- filter polygons using vaughn points; remove blank color boxes in legend, simplify legend title (include units in legend)
- make 2 close-up impages showing very successful zones, include labels of various small features using Vaughn's points;-Fig 1 zoom in so that training areas barely contained in extent, add graticule, add points to legend;send mxd of main image files to SKN & CB;<SKN & CB: edit draft, massage text.
Activities for 4/18:
adjustments to model: a) mask out Yellowstone Lake, b) widen coefficient ranges in MC for topographic variables, c) reduce study zone to reflect Watson et al. fig 3, d) run larger training areas to re-check coefficient ranges, e) mask crazy mountains;comparison ideas: let's compare to identified hot zones on YNP map rather than outlined mineralization zones
when plotting, do no-fill-color for pixels <0;{harass CB - send files to Sita for comparisons}
Activities for 4/10:
envi resampling across full park geometry (contact CB with issues);run residual model on full park;delete pixels under 1.5 and 2 stddev - do not rescale anomalies to 0;-> Finish first draft of paper
Activities for 3/13:
<confirm affiliations for Mono Map>;<continue to harass CB>;turn outline into first draft;cut "holocene" from scarp top on map;once CB model arrives, run and compare to ours
Activities for 3/6:
Compare possible false positives with Watson et al.;pick several images for GRC submission;double check intercept values from multivariate;write outline of GRC paper, including figures
<ask CB for output from Watson code;Mono Map: do "divided into" format for Trachydacite of Hot Springs Cove (and add "of" to Qhst);New map symbol for geomorphic features
for Qhst: convert map symbol to Qhstf, add word "flow" to end of name (like other flow units)
Activities for 2/24:
Confirm best-fit line type for each category in training;max and min coefficients for each category for Monte Carlo;run MC for zone in YNP, show figures filter residual at 1, 1.5, and 2 stddev (like Watson);change text numbers for MonoMap, export copy for SKN to see;
Activities for 2/6:
export Mono Map for Shannon;add wildcat drillhole (if we get coordinates);do new method for training areas;do 2 stdev filter on training zone results (looking for random distribution of "hot" pixels)
Activities for 1/23:
For each test area: raw vs. singular independent variables to find equation of best fit; make sure whether each ind. variable is best represented by line, power law, etc., for ALL training areas, work systematically from most to least important variable, using trend correction between variables; Mono Map updates
Activities for 1/16:
check univariate linear assumptions systematically, starting with most influential variables (esp: aspect),if nonlinear, update model, with corrected output (with subtracted filter), check for and label false positives, work on edits to P1, {Read/comment new draft, P2}
Activities for 1/9:
re-run residual filter using 1, 1.5 and 2 stdev of whole residual curve, label "white spot on top of mountain" in fig T3, put together series of images with limited explanatory text walking someone through entire process for one image, {SKN: edit P1}
Activities for 12/19:
check univariate linear assumptions systematically, starting with most influential variables, look for features in YNP that are relating to unexpected hot zones in residual; {SKN: look over code, reset permissions in parallel acct}
Activities for 11/21:
convert output to emittance, compare with Watson et al.,check temperatures in raw image, try adding huge buffer space on anomaly a;ready confirmed by Watson and black polygons, see if pixel ratio improves
Activities for 11/14:
send SKN Watson paper,5 color legend, centered on 0,big hot area in west-central park,look for snow melt map over anomalies for comparison, {contact RO for slides, draft thesis}
Activities for 10/31:
YNP: increase buffer size around known anomalies, increase number of colors in temp displays,overlay fault trace map,more hot zones,Can we explain our highs/lows relative to published anomalies, read/abstract another Mitchel paper on BLF/SRP,How were anomalies defined
Activities for 10/24:
Take on Yellowstone/BLF, now that websites are back up, read/abstract BLF paper (see Pollun thesis for ideas), {Sh: seriously, edit.}
Activities for 10/17:
take on Yellowstone ,take on BLF if time, read/abstract 2 papers: Yellowstone and BLF geology
{Sh: edits!}
Activities for 10/10:
contact CoSE IT for VPN key,set up VPN & login to SKN's desktop,try to log in to desktop, make personal folder on Desktop, send to Carrie, write/edit P2,figures, per discussion, {Sh: edit P1}
Activities for 9/24:
Add absolute diff residual mean value to table for training areas, rough draft figure of temperature difference per pixels as frequency curve, for 1 training area
send BLF Landsat to Carrie,continue edits, {Sh: omp, pi}
Activities for 9/17:
{Sh: continue editing P2; omp fix?, send SK the map comments}, edit P1, green section, run code for higher number of iterations, with timer, update short report,put together clean figures for P1 results/discussion sections, work on map comments
Activities for 9/10:
{Sh: edit paper 2 draft}, revise paper1 per comments, Mono Map edits as discussed via skype messages and documents, short report (annotated figures) on serial code output vs. Arc
Activities for 8/27:
Mono Map: add figure #'s, check that "approximate" locations for inset photos are represented on main map; Sh will finish shaded relief map annotations and send for embedding, resend Ch 2 (I'm so sorry I lost it!!!), {Sh: edit Paper 1, working on omp}
compare serial code output to Arc output 1) using Arc coefficients and 2) using random numbers. If residuals = 0, convert "real" numbers to double precision real
Activities for 8/20:
Arc best-fit equation plus residual by training area (send to Sh), {Sh continues to debug, edit}, edit A-B =C figure, edit draft paper, make list of key figures w/ short descriptions
Activities for 8/13:
{Sh: work on debugging code; edit text}
send Sh the Arc-based residuals for each training area, take a look at remote sensing journals looking for one with relatively short articles, check figure guidelines and convert existing figures to match,read another article, if edited text received, work on edits
Activities for 8/6:
Debug OMP. (why slower? pull coefficient generation out of omp loop control), read Wagner and Fortin (2005), emphasis on methods, approaches (don't worry about ecology), plot map of residuals for training areas using best-fit solutions; compile table of best-fit solutions for training zones, first draft of first paper (first: pick your journal)
Activities for 7/3:
Debug residual problem in code, Debug timer,Mono Map move, read Wagner and Fortin (2005), paper from Carrie (geothermal Yellowstone RS)
[Shannon: edit abstracts, send]
Activities for 6/25:
Comment and indent code, including title box explaining code and naming variables, write comments in code showing where splits should occur for MPI and openMP
time program for first pass (10 iterations) on each training area, re-contact MM about prospectus, start methods for Ch 4{Sh: read/edit methods}, read 2 papers (Monte Carlo for multivariate), Mono Map image moving
Activities for 6/18:
rough draft of multivariate code using random numbers, contact Mike about signing prospectus, send copy of prospectus to GFR & Carrie, work on methods text for thesis (2 chapters!), separate references by chapter
Activities for 6/4:
figure out standard deviation for multivariate solutions for coefficients - pick overall high and overall low (mean +/- stddev) for each category, work on OMP debugging for race conditions, email JW regarding his timeline for comments or signature, map how to upgrade code to multivariate
Activities for 5/17:
Updates to openMP version of code, test with large datasets, incorporate edits as needed (contact JW if necessary), Mono Map, Prepare graphs or tables to show the different coefficients calculated for August in different years and diff training zones
Activities for 4/25:
Work on serial version of Monte Carlo code univariate, compare results to Arc output, Mono Map graticule, continue working on CB's comments
Activities for 4/9:
Standard deviation of residuals relative to expected anomaly, make residual maps colored based on standard deviation, clean multivariate coefficients table, include the mean and std dev, Colloquium talk draft (22 min, aim for 22-28 slides), test model on Yellowstone, send edited prospectus to committee with letter describing changes)
{For future: UTM graticule on Mono Map, confirm whether text stays messed up when opened on PC}
Activities for 4/4:
email Carrie and JW about prospectus comments, subtract residual maps from the two temporal image groups, start on multivariate coefficients table, identify anticipated anomaly magnitudes from literature, compared to existing residuals
Activities for 3/21:
meet with Carrie about rock types, mineral detection, run training for second date, confirm continuing residual, incorporate MM's comments, Monte Carlo
Activities for 2/28:
prepare presentation for committee meeting, clip rock maps, rock group comparisons, mono map: 5 line changes/deletions
Activities for 2/28:
email committee to set up meeting, Side-by-side images of aerial photography and shaded relief of training areas: for appendices, commmittee meeting, finish Ch 2 edits
work on comm mtg presentation, geologic maps vs. lithology signal
Activities for 2/14:
error catching: normalized elevation in area, hydrologic layer, anything else?, remove aspect and slope, set a meeting day/time for committee, Ch 2 edits, Side-by-side images of aerial photography and shaded relief of training areas: for appendices, comm. meeting.
Activities for 2/7:
small area generalized descriptions, look at possible variables to add, % remainder examples for committee, (Sh: prospectus! ASAP!))
Activities for 1/31:
Investigate anomaly sources, match ranges for R^2 maps, finish up next draft of prospectus (goal: send to committee, invite to current results meeting), start training area descriptions, update R^2 table for dropping NDWI, and include the complete training zone
Activities for 1/14:
(Sh: send ref letters), test effects of dropping NDVI vs NDWI for multicollinearity; impact on NDBSI?, Impact of dropping slope?, thermal inertia images for sale?
start report-style summary of regression results tables, figures, very short text summaries
Activities for 1/7:
(Sh: edit prospectus), univariate at other training locations; table summarizing significance, targeted multivariate analyses, training zones
Activities for 12/17:
univariate analyses of training zones, looking for significance, soil moisture, rock/mineral, insolation time, EVI, update prospectus
Activities for 12/10:
univariate analyses of training zones, looking for significance, as much as possible, (Sh - finish editing prospectus, send to Sita)
Activities for 11/16:
continue on last week's papers (send to Shannon; write analyses & send to Sh), edit prospectus, metadata table
Activities for 11/5:
find 2 new papers (send to Shannon; write abstracts & send to Sh), finish metadata table, (Shannon - read/edit prospectus)
Activities for 10/29:
finish editing prospectus (v2), find 2 new papers (send to Shannon; write abstracts & send to Sh), collect imagery, - put together a table of metadata
Activities for 10/22:
(Shannon - edit prospectus, hard copy), find 2 new papers (send to Shannon; write abstracts & send to Sh), edit prospectus, Check writing webinars by Schimel (http://vimeo.com/17432903)
Activities for 10/8:
(Shannon - edit prospectus, hard copy), edits prospectus refs, find 2 new papers (send to Shannon), if reviews back in time, edit
Activities for 10/1:
5 pages text for prospectus, find 2 new papers (send to Shannon)
Activities for 9/24:
Find multivariate regression in Arc paper, send to Shannon, demonstrate regression with any 1 independent variables, make list of high-risk multicollinearity X
multivariate regression in mountainous US (any topic), start very rough outline of thesis
Activities for 9/17:
Download 10 m elevation data, create polygons of clean areas - avoid WY, Read Ch 8, do Q5, For list of independent variables, identify source data for each
Activities for 9/9:
read Ch 7, do problem 3, Look for DEM for area, Find out about geodatabase for Yellowstone (waiting on response), (Shannon:find a problem in Ch 7, read Aster paper)
Read Maselli (2002) and Franco et al. (2000) to discuss next week, Add goals text
Activities for 9/5:
Find layers: for test and training datasets (E&N of Yellowstone site, ESRP), (Shannon: find spatial stat book from Sarah, email John Welhan about hot spring maps)
Read multivariate analysis chapter [EDIT: I want to change this. Let's read Ch 6 instead (Intro to Regression) and solve problem 8 (p. 123)]
Identify prospective committee members