Dwarf Fortress Bug Tracker - Dwarf Fortress
View Issue Details
0008781Dwarf FortressWeatherpublic2015-02-02 11:232017-08-03 19:06
lowtrivialhave not tried
0008781: Multiple biomes in air above surface; leads to odd weather like raining blood above "good" biome.
My current fortress straddles a good biome and a neutral biome. However, I frequently receive alerts that it is "raining human blood." However, it seems to evaporate before it hits the surface, although it does seem to collect in the tree tops.
Settle on a line between good and neutral biomes.
There is an evil biome relatively close to the northwest, although it is not on the local map.
No tags attached.
parent of 0009344new  Eerie Murk-covered Raven Thralls appearing on non-evil embark 
has duplicate 0008724resolved Loci Evil rain in a neutral-savage biome 
has duplicate 0009477resolved Dwarfu Biomes different in air as opposed to ground's biome type. Weather gets weird because of it. 
has duplicate 0009476resolved Loci Evil weather in non-evil part of the map. 
has duplicate 0007399resolved Loci Deadly contaminant in fortress with no evil biomes 
Not all the children of this issue are yet resolved or closed.
Issue History
2015-02-02 11:23BSRoweNew Issue
2015-02-02 13:57DetrosNote Added: 0032123
2015-02-02 17:12BSRoweNote Added: 0032124
2015-12-14 08:49PatrikLundellNote Added: 0033879
2015-12-14 10:56DetrosNote Added: 0033883
2016-07-25 12:24viperstingNote Added: 0035702
2016-07-25 12:41viperstingNote Added: 0035703
2016-07-28 05:41DeKaFuNote Added: 0035714
2016-10-03 11:31PatrikLundellNote Added: 0035931
2016-10-03 18:06LociNote Added: 0035935
2016-10-03 18:08LociRelationship addedhas duplicate 0008724
2016-10-03 18:12LociSummaryRaining blood in "good" biome. => Multiple biomes in air above surface; leads to odd weather like raining blood above "good" biome.
2016-10-03 18:13LociRelationship addedparent of 0009344
2016-10-03 18:16LociRelationship addedhas duplicate 0009477
2016-10-03 18:17LociRelationship addedhas duplicate 0009476
2016-10-03 18:18LociNote Added: 0035937
2016-10-03 18:19LociNote Added: 0035938
2016-10-03 18:19LociAssigned To => Loci
2016-10-03 18:19LociStatusnew => confirmed
2016-10-03 18:22LociNote Added: 0035939
2016-10-03 19:58LociRelationship addedhas duplicate 0007399
2016-10-03 20:01LociNote Added: 0035942
2016-10-04 23:22PatrikLundellNote Edited: 0035931bug_revision_view_page.php?bugnote_id=0035931#r14487
2016-10-05 03:39PatrikLundellNote Added: 0035945
2016-10-10 17:164maskwolfIssue Monitored: 4maskwolf
2017-06-08 19:34HuntthetrollIssue Monitored: Huntthetroll
2017-08-03 19:06LociNote Added: 0036687

2015-02-02 13:57   
Can you upload your save folder that shows this behaviour to http://dffd.bay12games.com/ [^] and post a link to it here, please? Also, version for which OS are you using? Are there any mods used or is it just vanilla DF 0.40.24? Is this bloody rain happening on the whole embark site or just a part of this map?
2015-02-02 17:12   
I'm on a PC, version 40.24, no mods, just a graphics set (Ironhand). The blood was localized to a small area, BUT since I saved and reopened, it has since stopped. I'll keep an eye out to see if it comes back, but for the time being, it appears to no longer be a problem.
2015-12-14 08:49   
http://dffd.bay12games.com/file.php?id=11477 [^]

This save contains a non evil embark subjected to evil rain that doesn't reach the ground, but is collected in graphically invisible puddles (detectable by 'l'ook) 1 a 2 z levels above ground, as well as on my dog box guard tower. It seems these puddles appear in line with and north of the dog box, but I might just have been unlucky with my panning.

The embark did also get 3 waves of undead flyers in the year before I abandoned it (these flyers do not exactly encourage visitors, which I wanted to try out).
Ground based critters entering were normal for the embark, though.

The save is from 0.42.01 using LNP r02 with the Phoebus tile set. My computer is a PC with Windows 8.1.
2015-12-14 10:56   
Are there any evil regions near your embark site?
2016-07-25 12:24   
I might have an idea of what is causing this bug:

http://dffd.bay12games.com/file.php?id=12297 [^] [v0.43.05] (1)

http://dffd.bay12games.com/file.php?id=12298 [^] [v0.42.06] (2)

In save (1), the fortress is located in a neutral biome, in the roof of the fort (1 z-level above ground) evil clouds are generted

Save (2) is similar, evil rain falls in the surface, but no in the roof. The roof will also spawn evil clouds which do not correspond to the biome.

Both Evil clouds are generated in biomes adjacent to the fort's biome [(1): Southern evil mountain. (2): Northwestern evil forest]; these biomes are located in the same local map as the fortresses (16x16 embark tiles). The rain which spawns in the save of PatrikLundell is generated in a evil forest to the east located also in the same local map.

The biomes which spawns the evil weather have a heigher altitude (at least 1-z level higher) than the surface of the fortress embarks.

It appears that the game will assignate evil weather to different z-levels according to each biome altitude, but instead of applying it to every respective biome, it will apply it to the whole local map in the embark screen. Generating biomes with misplaced evil weather.

Ex: Biome 1 [z-levels 0 to 1] is neutral, Biome 2 generates evil clouds [z-levels 2 to 5]. As both are located in the same Local map, embarking in biome 1 would show evil clouds in z-levels form 2 to 5.
2016-07-25 12:41   
After placing a refuse stockpile over the roof in save (1), water buffalo hair started to reanimate. This is also a characteristic of the mountains in the south. This means that the reanimation grounds can also be misplaced due to this bug.
2016-07-28 05:41   
I've been keeping track of this bug since I believe it's likely related to my issue 0009344 (contagious husked birds in neutral biome).

Other related/duplicate issues to this one:
2016-10-03 11:31   
(edited on: 2016-10-04 23:22)
Some additional input based on DFHack script making for 0.43.03:
The air over embarks are frequently a messy patchworks of biomes, some of which are present at the ground level as well, and some that are not, although you tend not to detect that unless you get evil rain, snow, or inappropriate fliers, or alternatively set up farm plots above ground what won't support growing of the plants they ought to. The biomes can be different at different Z levels. In some cases the offset data elements in the data type element DFHack calls block.region_offset is way out of the 0 - 8 range they're supposed to have. However, the patchwork biomes can be present even when these offsets are within bounds, so they're not the cause, but probably merely a symptom of something that's off.
Edit: Another observation: Embarking twice at exactly the same location using copies of the same pre embark world save can result in a different set of anomalous air biomes in different locations.

2016-10-03 18:06   
v0.40.23: Teres_Draconis posted a save in 0008724:

http://dffd.wimbli.com/file.php?id=10430 [^]
2016-10-03 18:18   
v0.42.04: Di posted a save in 0009476:

http://dffd.bay12games.com/file.php?id=11659 [^]
2016-10-03 18:19   
v0.42.04: Walkabout posted a save in 0009477:

http://dffd.bay12games.com/file.php?id=11660 [^]
2016-10-03 18:22   
@DeKaFu: Thanks! I've consolidated those reports here.
2016-10-03 20:01   
v0.40.03: Di posted a save in 0007399:

http://dffd.wimbli.com/file.php?id=9022 [^]

v0.40.3: Krandall posted a save in 0007399:

http://dffd.wimbli.com/file.php?id=9023 [^]
2016-10-05 03:39   
Yet another piece of the puzzle. I've debated whether it should be a separate report, but decided it's too closely related, as a user would be hard put to separate the two, so error reports would report very nearly the same thing and the underlying technical reasons are closely related as well.

The air space above an embark frequently has a different biome from the embark itself. In addition to that, there can also be biomes sheared into this, which is covered by my previous entry.

I suspect the reason the main airspace biome frequently differs from the biomes at the surface is that the array substructure DFHack calls block.region_offset defaults to 0 for blocks in the air. However, 0 means "the world map tile to the NW of the one referenced by this block" (and that reference should be the block's own world map tile), rather than 4 (no change). With a 4 rather than 0 the airspace would default to the biome of the world map tile (which might still not match the embark, especially if the embark is not in the center of the world map tile).

It can also be noted that the offset overflow handling seems to result in an effective usage of an offset value of 4 (i.e. no adjustment).
2017-08-03 19:06   
PatrikLundell posted a DFHack workaround on the forums:

http://www.bay12forums.com/smf/index.php?topic=165081.msg7526344#msg7526344 [^]