Blank Climate Projections

[Matt]

Hi,

I'm relatively new to MAXENT and am trying to model the impact of
climate change on bats in the UK at the moment, but have come a bit
stuck when using the UKCIP02 outputs. I'm using four climatic fields
for projecting which have all been converted to ASCII and the file
names are identical to the current environmental fields and have
exactly the same extent and cell size. The current distribution output
looks great but the projection appears virtually blank?

I'm not sure why this is happening as I'm expecting this species to
extend it's current range northwards. Does anyone have any ideas?

Cheers,

Matt.

[Andrew Davies]

Hi Matt,

Are the ASCII outputs looking blank in ArcGIS, with huge numbers for
the range? If so, then you can convert your ASCII output to a FLOAT
raster (see other posts on the groups) or just run your ASCII through
RasterCalculator to rebuild the histogram and it should display your
output.

Best,
Andy

[Abhi]

Hi all,

My name is Abhishek. I am relatively new in ecological niche modeling.
I have a question regarding climate change projection in Maxent
(3.3.1). Through the prediction at current climate situation looks
very good.. the projections at future scenarios yielded blank map. I
am using future climate data downloaded from Worldclim (2.5 arc
minute resolution), and the bioclims were calculated through Diva.
Converting the projected ASCII into FLOAT raster didn't help either.
Though i believe the both future and current climate variables are at
the same spatial resolution and there were no error while MAxEnt was
running, I am still pasting below the headers of the one each of the
future and current climate variables, in case something is mixed up
there. If this problem is too basic for all of you please excuse me, I
am certainly not a pro in this field. It will help me a lot, if some
body please can help me out in this regard.

bio_1 (2020, A2a)
......................
ncols 6736
nrows 2006
xllcorner -122.958333
yllcorner -40.541667
cellsize 0.041666666567696
NODATA_value -9999

bio_1(current)
..............................
ncols 6736
nrows 2006
xllcorner -122.958330358378
yllcorner -40.5416656518353
cellsize 0.041666668839753
NODATA_value -9999

Thanks, Abhishek

On Aug 11, 5:06 am, Andrew Davies <andy.marine...@googlemail.com>
wrote:

> Hi Matt,
>
> Are the ASCII outputs looking blank in ArcGIS, with huge numbers for
> the range? If so, then you can convert your ASCII output to a FLOAT
> raster (see other posts on the groups) or just run your ASCII through
> RasterCalculator to rebuild the histogram and it should display your
> output.
>
> Best,
> Andy
>
> On Aug 2, 4:57 pm, Matt <shotty_...@hotmail.com> wrote:
>
>
>
> > Hi,
>
> > I'm relatively new to MAXENT and am trying to model the impact of
> > climate change on bats in the UK at the moment, but have come a bit
> > stuck when using the UKCIP02 outputs. I'm using four climatic fields
> > for projecting which have all been converted to ASCII and the file
> > names are identical to the current environmental fields and have
> > exactly the same extent and cell size. The current distribution output

> > looks great but theprojectionappears virtually blank?

[Milton Cezar Ribeiro]

Hi there,

 

Try to copy one header to all files, and try again :-)

All information need to be EXACTLY the same to maxent run fine.

 

cheers

 

milton

2009/8/26 Abhi <abtr...@gmail.com>

[Lucero Sevillano]

Hi Matt,

I am new to Maxent myself, and I am modeling the impact of climate change on an invasive tree. The current distribution looks fine, but in my projection distribution I get a blue map. I don't know if when you said you were getting a blank map you meant literally blank, or if, like me, your probabilities were extremely low. The picture comparing environmental similarity between future and current conditions (multivariate similarity surface, MESS) is almost all red, meaning that almost all values in the future are outside the current range. 

How did you solve your problem? I wonder if somehow my future variables are wrong? I obtained the 19 bioclimatic variables for the future (worldclim, IPCC4) following the protocols of Ramirez and Blanco-Cabrera (2009), using ArcGis and DIVA.

Any help will be much appreciated!

Cheers,

Lucero  

[Alexander Gamisch]

Hi Lucero,

i got a similar experience when projecting the current distribution of my species into past climate models (LGM) using the layers provided by the worldclim site. Therefore i think it should not be the problem of your future variables if you prepared them after the protocols of Ramirez and Blanco-Cabrera (2009).

However i would also be very curious what the more experienced Maxent user say about this problem.

I am happy for any clue or opinion!

Kind regards!

Alex

2012/11/22 Lucero Sevillano <lsevi...@cieco.unam.mx>

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To view this discussion on the web visit https://groups.google.com/d/msg/maxent/-/6FuUgtOecoIJ.
To post to this group, send email to max...@googlegroups.com.
To unsubscribe from this group, send email to maxent+un...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/maxent?hl=en.

[swammy50]

Are the units of your future variables the same as current? The red mess is a clue. Use GIS or R to look at the distribution of the climate variables to check if they make sense.

Sam

[Lucero Sevillano]

Ok, I just reviewed the .asc files in a text editor, and found out that for the current variables, there are some extreme values such as 55537 in areas where there are no data (sea), and for the same variable in the future these same areas have negative values, such as -29997. Besides that, it looks like the values of the actual study areas (Central Mexico) make sense. For instance the precipitation range for current conditions is within the range for future conditions.

Now, should I just replace those values in both sets of variables (current and future) with the same value, for example -9999? I am worried that that would mess up my variables :S

Thanks everybody!!

Lucero

[Lucero Sevillano]

Dear group, 

I also found out that temperature for current conditions is in units of ºC TIMES 10 ... (http://www.worldclim.org/bioclim-aml), and future values are not multiplied by 10. That explains, at least partially my problem! 

Maybe some of you knew this ... I´ll try again after correcting variables related to temperature.

Thanks!

Lucero

[swammy50]

Make sure your precip variables are in consistent as well. Also your data should have the same no data values, sounds like you may have a problem with that to.

Sam

To view this discussion on the web visit https://groups.google.com/d/msg/maxent/-/5HWF-MZ-wuYJ.

[Rita Afonso]

Dear group,

I know it has been a while... but I'm having the exact same problem. In the projection I made for LGM, the probabilities are extremely low.

I'm not being able to understand why. I would really aprecciate if you have some suggestions.

Thank you very much,

Rita