Back-of-envelope estimation of total stem count in NSSF

[hrlai]

Using the H3 report Table 3-5, the upper estimate is ~65 stems/ha, simply scaling up to 750 ha gives ~48750 stems.

[hrlai]

471334 stems at year 47
100 year should finish running tomorrow noon, we'll see how many at year 100...

[hrlai]

Ignore the above, print message was coded incorrectly

[hrlai]

51687 stems at year 60   
53591 stems at year 61   
56678 stems at year 62   
59311 stems at year 63   
61673 stems at year 64   
64727 stems at year 65

[hrlai]

And now

77711 stems at year 70   
81084 stems at year 71   
83992 stems at year 72   
87097 stems at year 73 

Given these numbers, the static "all other species" number at init.aos.n <- 10000 becomes increasingly unrealistic too... I wonder if the static AOS density also contributes to Prunus not slowing down in population growth...

[hrlai]

Note-to-self: the back-of-envelope calculation from H3 report does not include stems <1-cm DBH (?)

[wengngai]

I finally did a proper calculation of the number of stems present in all of NSSF. Here's the summary:
trees (DBH > 5cm): 0.135 stems per m^2
saplings (5 cm > DBH > 1cm): 0.420 stems per m^2
seedlings (DBH < 1cm): 4.426 stems per m^2*

• There is a bit of selection in seedling plot location though. so this number actually may need to be almost halved. Because we had rather strict criteria for where seedling plots could be located (e.g., not under palms, etc.) so that the chance of sampling seedlings would be higher

Summed up across the 75-ha catchment (I used the same polygon used to clip the IBM base maps):
trees: 1,020,479
saplings: 3,171,812
seedlings: 33,424,301
total: ~37.6 million

clearly my initial population size guesses were very close. just about 4000x less than reality.
Also, I have uploaded an excel file (not csv) in the /data folder with summary statistics of basal areas, stems, etc. of adult trees (DBH > 5cm) by species, just in case you need any quick info with which compare simulations.

[hrlai]

Haha. But you were guessing for the smaller spatial extent last time. 😊

It’s good to know your guesstimate is way larger then my more-naive guess, and is much higher than the simulation.

The thing is that if the seedling number can really go to this high, then the simulation will surely still be very computationally challenging. But let’s see if we could come up with some clever ways...