Poor fire return interval fits suggest new data cleaning steps

[dschwilk]

Or fire regime fitting ended up being quite poor (see https://github.com/Fireandplants/modis_fire/issues/9). So we are going back and revisiting some cleaning decisions.

We suspect that human modified landscapes and tree plantations may be messing things up @SallyArchibald suggested 2 steps:

1) use Human Impact Index

"After much research it appears there is no global land cover product that will show us plantations. I suggest not using a land cover product: just cleaning for the Human Impact Index >30, with perhaps some additional cleaning of Pinus from southern hemisphere Africa etc."

http://sedac.ciesin.columbia.edu/data/set/wildareas-v2-human-influence-index-geographic

2) Clean for known plantation observations (Eucalyptus outside Australia, Pinus in southern Africa)

[dmcglinn]

@SallyArchibald has also suggested filtering for non-cultivated land using this MODIS product:

https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/mcd12q1

@SallyArchibald I found this distribution of that dataset:

https://e4ftl01.cr.usgs.gov/MOTA/MCD12Q1.051/

How difficult is it to assemble all these files into a single raster layer? Or has someone already do this for us? Also I already have the HII dataset ready to go.

[SallyArchibald]

Hi,

I am happy to create a MODIS product which is coded 1/0 for transformed vs untransformed land. Would that do?

Will get onto it today and get it to you guys in a day or two

Sally

---

Sally Archibald Assoc Prof, Animal Plant and Environmental Sciences, University of the Witwatersrand +27 11 7176420 +27 83 966 2544 sally.archibald@wits.ac.za

On 24 April 2017 at 20:48, Dan McGlinn notifications@github.com wrote:

@SallyArchibald https://github.com/SallyArchibald has also suggested filtering for non-cultivated land using this MODIS product:

https://lpdaac.usgs.gov/dataset_discovery/modis/modis_ products_table/mcd12q1

@SallyArchibald https://github.com/SallyArchibald I found this distribution of that dataset:

https://e4ftl01.cr.usgs.gov/MOTA/MCD12Q1.051/

How difficult is it to assemble all these files into a single raster layer? Or has someone already do this for us? Also I already have the HII dataset ready to go.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Fireandplants/plant_gbif/issues/22#issuecomment-296787420, or mute the thread https://github.com/notifications/unsubscribe-auth/AGjWCgu3sqHFiCqGU07ivDErdYOE8DRHks5rzO6EgaJpZM4M6KGT .

[dmcglinn]

That would be super helpful. Thanks! Dan

On Fri, Apr 28, 2017 at 10:17 AM, Sally Archibald notifications@github.com wrote:

Hi,

I am happy to create a MODIS product which is coded 1/0 for transformed vs untransformed land. Would that do?

Will get onto it today and get it to you guys in a day or two

Sally

---

Sally Archibald Assoc Prof, Animal Plant and Environmental Sciences, University of the Witwatersrand +27 11 7176420 <+27%2011%20717%206420> +27 83 966 2544 <+27%2083%20966%202544> sally.archibald@wits.ac.za

On 24 April 2017 at 20:48, Dan McGlinn notifications@github.com wrote:

@SallyArchibald https://github.com/SallyArchibald has also suggested filtering for non-cultivated land using this MODIS product:

https://lpdaac.usgs.gov/dataset_discovery/modis/modis_ products_table/mcd12q1

@SallyArchibald https://github.com/SallyArchibald I found this distribution of that dataset:

https://e4ftl01.cr.usgs.gov/MOTA/MCD12Q1.051/

How difficult is it to assemble all these files into a single raster layer? Or has someone already do this for us? Also I already have the HII dataset ready to go.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Fireandplants/plant_gbif/issues/22#issuecomment- 296787420, or mute the thread https://github.com/notifications/unsubscribe-auth/ AGjWCgu3sqHFiCqGU07ivDErdYOE8DRHks5rzO6EgaJpZM4M6KGT .

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/Fireandplants/plant_gbif/issues/22#issuecomment-298009942, or mute the thread https://github.com/notifications/unsubscribe-auth/ABIp7aPJWKRRoMM-WWjY1ZCMG1yhGTHeks5r0fTegaJpZM4M6KGT .

-- Daniel J. McGlinn, PhD Assistant Professor College of Charleston Department of Biology Harbor Walk West, rm 203 360 Concord Street Charleston, SC 29401 http://mcglinn.web.unc.edu/ office: 843-953-0190 cell: 405-612-1780