Posts Tagged ‘land mask’

A real land mask

As mentioned in the previous post, I now have a real land mask from ISLSCP Initiative II.

The format of the files in refreshingly simple. 180 rows with 360 space separated numbers. It takes nothing more than a few dozen lines of Python to convert this into the 8000 cell format that ccc-gistemp uses for its step5mask.

Here’s the 8000 GISTEMP cells that contain any land (according to the ISLSCP file). Drawn on a Plate Carrée projection:

Even at this blobby scale there are some suspicious features (missing). Where is Ascension or Cocos Islands? I don’t think it’s an artefact of my processing, because they don’t appear on the ISLSCP file either; that’s probably a bug. So accepting that their might be some minor issues with islands and so on, we can go on to do a run of ccc-gistemp using this land mask:

As mentioned in the previous article, restricting to land de-emphasises coastal stations which generally warm less slowly. So the restricted version of the analysis has a stronger warming trend.

Many of the partial land cells will have ocean data and will not have a nearby land station to count as a land cell in the usual GISTEMP analysis. So we find that excluding cells that have ocean data (and no nearby land station) results in an even stronger trend. The red graph is reprised from the previous article:

(pedants should note that in the above graph “cells with ocean data” means “cells with ocean data and where the nearest land station to the cell’s centre is more than 100 km away”)

There are countless variations one could try such as weighting cells by their coverage, or using a threshold of 50% land cover. As far as an estimate of land temperatures go, such variations amount to interpolating between those two curves on the previous chart. Roughly.


[update: 2010-08-26: the “no arctic” masks were wrong, fixed now]

I added a land mask feature to ccc-gistemp. The land mask is used In Step 5, where land and ocean series are combined into zonal averages. For each cell, ocean data is used unless: there is very little ocean data (fewer than 240 months); or, there is a land station within 100 km. With the new land mask feature an external file (input/step5mask which is the same format as the newly output work/step5mask) specifies whether to use land or ocean data for each cell.

This enables me to do something new: a run of ccc-gistemp using only land-data but restricted to locations where there is no ocean data:

I don’t actually have a true land mask to hand (a list of cells that cover the Earth’s land surface), but I can do a normal run of ccc-gistemp that uses both land and ocean data and use the generated mask. That mask tells me which cells have no ocean data (and therefore land data is used, if there is any).

That’s the mask I use above. This reduces the effects of the 1200 km smoothing radius, by preventing land series from being interpolated over the ocean (unless there is no ocean data). This mask has 3077 cells (8000 in total), so clearly some ocean cells are still using land data. It looks like this:

Note that black is not land, but it is where land data is used if there is any. Note the arctic ocean has no ocean data so land data (generally interpolated) is used. Also some of southern ocean will pick up land data.

The fact that the land series when restricted to exclude cells with ocean data shows greater warming can be explained by considering coastal stations. In the usual ccc-gistemp land-only analysis every station has an influence that extends out to a disc of radius 1200 km. Coastal stations will have a disc of influence that extends into the ocean. When excluding cells that have ocean data that means that a coastal station has its disc chopped in half (roughly), and an island station is almost entirely excluded. The overall effect is to weight stations in the continental interior more strongly. And warming is greater in the continent interior.

[edit: Chad also shows a version of the same effect (scroll down the very long post until you reach the red and black graphs): de-emphasising coastal stations increases trend]

We can show the effect of excluding Arctic and Antarctic zones (everything north of 60N and south of 60S is excluded):

Only a small effect. The broken mask version had an even smaller effect, but I’m still a little surprised by how small this difference is.

Again we can exclude cells with ocean data, and again this shows increased warming. This more than cancels out the cooling by excluding the high latitude zones:

Anyone got a real land mask? [edit: Yes, Steven Mosher pointed me at one; so expect a post using a real land mask soon]