Maximum cumulative suitability value less than 1

[Esass]

Hello,

I am trying to use Maxent to predict the habitat suitability for a species currently found in two distinct populations. When I train the model on one population and test it on the other, the highest suitability value given to any pixel across the area is 0.6. When the model is trained on the second population and tested on the first, the highest suitability value is about 0.9. Shouldn’t the highest cumulative values be 1 in both cases? And why should one direction be so different from the other? Any advice or direction would be great. 

Thank you,

Emma 

[Megan S]

I have the same question. I am changing the variables input to my model and in some cases my maximum suitability value is 0.9 and others it is as low as 0.7. What does that say about the model? They have the same AUC values.