Hi LomasDelGuijarro,

Be careful realigning roads in Lusaka based on satellite imagery. There has been a great deal of road construction recently that does not appear in the imagery. Some of the new roads are from GPS tracks.

Further to this all rivers in Zambia can be classified as seasonal as we have only 2 real seasons a wet from November to March and dry from April to October although peak flows in the Zambezi are usually April May, something to consider if you intend visiting the Victoria Falls. some rivers I would classify as both seasonal and intermittent as even in the wet season they will only really flow when when rain falls but may start to run morre regularly once the dambos are waterlogged and start to drain, note many river systems start in dambos. In a 'normal' wet season this will be around February but we have not had one of those for some time. In the lower Zambezi Valley the rivers rush out of the hills in flash floods and then will not run until the next rainfall although there will be water below the river bed, your classic elephant drinking holes.

Just some local knowledge for you.

Hi I have been slowly working my way down the rivers in this area. Takes some time as the area is very flat and rivers have many bends in them. They also run in what we call dambos in this part of the world. A dambo is a general term for a seasonal wetland and in the dry season can be a dry grassland and will burn fiercely hence the blackened areas that some mappers mistake for water. The western part of Zambia is also very sandy so river courses can change regularly. The sand can be seen as the white or pale areas in the rivers especially the Zambezi and the white rims around some of the dambos. If you look up dambo on Wikipedia you will see what an important feature it is in Zambia

Imagery over Zambia tends to be hit and miss and not updated regularly. I have had Apple Mappers moving through towns in Zambia in the past and making changes to features and roads that have been mapped by local mappers but do not appear in imagery. This can be a bit annoying. Also the mappers tend to follow each other nudging nodes and not taking into account that imagery may be offset. This has resulted in roads I know to be straight having slight kinks in them. I am not against people mapping in Zambia just sometimes you need to have a degree of local knowledge. I would never try map in the USA as I would not know what I am looking at in the imagery.

Lake Kariba fluctuates in level quite substantially so any edits you make now are likely to be inaccurate as imagery tends to be old. in fact at present Lake Kariba is close to historic lows.

Don't get too excited about being precise with the borders and features of Lake Kariba

This lake fluctuates in level quite substantially and the imagery is likely to be very old. Currently Kariba is close to historic lows so there are features visible that are usually below water level. Please inform Apple mappers of this fact

Hi,

Surely tagging this as a nature reserve is a better option, leisure:park implies a park within a town or city with benches flower beds etc.

Hi,

I have noticed you have been working on the Zambezi, just a note when working with multipolygons such as the Zambezi any islands are tagged with an inner role. If you imagine the relation being a layer with the outer border being tagged with outer and any island or feature within the bounds of the relation being a 'cut out' and tagged inner. In other words you 'cut' the island out of the 'Zambezi' layer leaving a hole for the island to show through. I have corrected the sand islands in the changeset perhaps have a look to get the idea.

Another point to remember is these islands may no longer exist as the imagery is usually outdated and the very nature of the islands means they change with the seasons as anybody who visits the Zambezi regularly will have seen.

Cheers

Hi mspray10,

I think some Mspray users are using a JOSM plug-in , possibly to square off buildings, which also squares off nearby ways and hence moves tracks or highways and sometimes breaks nearby relations. I have reverted some of the affected ways, not the buildings, but it would be better to prevent this rather than go around finding and fixing them. I have created many wetland relations and found some to have been affected this way. I did once use a building generalization plug-in as many buildings have not been mapped square. The problem was the plug-in had some strange results so I stopped using it. I have also come across errors in remote parts of Zambia where, possibly, a user has selected all the buildings in an area and squared them off automatically, possibly using the 'Q' key. this works well where generally buildings are square but not so well where there are circular huts, they end up as a straight line. In Zambian villages now days the residential houses may be rectangular but the grain stores and outside 'kitchens' will be circular thatched huts.

Just thought I would bring this to your attention.

Dave

Hi these areas you have mapped as wetland within Lake Bangweulu are actually areas of floating vegetation and or algae and as such will probably move with the prevailing winds so best to ignore them. The prevailing winds in Zambia vary with the seasons, coming from the east during our dry season May to November and from the West during the rains. You can sometimes see this in satellite imagery of different time periods when you zoom in and out on an area. In some wetlands in Zambia you can see the effect in open patches of water that have a strange cluster of circular patterns towards one end. These are the floating islands that have been pushed together by the wind. To see this in JOSM select the Bounding Box tab in the Download Dialogue, enter the following link osm.org/#map=17/-9.584357050831828/28.661141249315964 in the large URL selection area and download the data in this area. Switch between Bing and Digital Globe Premium imagery and see the islands move from one end of the open water to the other. In some other places the islands are almost perfectly circular as they slowly revolve with the effect of the wind. In open water further north, at coo-ords -9.5404944 28.6591458, you can see an island that has revolved slightly when you switch between images.

Hi I think you are using the 'Q' key in JOSM to square off buildings after making a multiple selection of buildings in an area. This works if all buildings are supposed to be square or rectangular but turns circular huts into lines with intersecting points. As the areas are in rural Zambia there is a good chance that many of the buildings will be huts, particularly if they are 'kitchens'

Be careful when squaring buildings using the 'Q' key in JSOM. It will turn circular huts into straight lines with intersecting points

You are messing up relations and squaring them off. Please stop you are destroying hours of work

By deleting some ways you are breaking existing relations. In this case the wetland bordering Lake Mweru. I have replaced the deleted way.

Hi,

There is no need to tag road restrictions where basically everything is allowed i.e access=yes, horse=yes etc as these are assumed. Also it is assumed that the road will have 2 lanes unless it is possibly a oneway with only say one lane or a dual carriageway with each carriageway having a number of lanes . In the case of a 'close' that has a dead end you can add the tag noexit=yes to indicate it has no through connection to another road. Where OpenStreetMap data is used for navigation devices it will indicate to the device the road is a dead end. Note it is noexit=yes not noexit=no, i.e. 'yes there is no exit to this road'.

Also in the case of relations a single highway should not be added as an inner section (the inner section of a relation should be closed, i.e. form a ring). In the case of the residential area relation an inner section or area, if there was one, would indicate that the area was not part of the relation i.e. a 'cut out' or 'hole' section. An easier way to understand this is in the case of a wooded area with an open patch of grassland in the centre. You can draw a line around the outer boundary of the wooded area then draw a line around the grassland area and tag it as natural=grassland. You then create a multipolygon relation which you tag natural=wood, the outer boundary, around the wood, is added to the relation with the role 'outer' and the boundary around the grassland is added with the role 'inner'. In effect this will create a wooded area with a 'hole' for the grassland. Do not tag the outer boundary as the tagging is achieved by tagging the whole relation. See this link for a more detailed explaination osm.wiki/Relation

I presume you have not randomly chosen the name for this road and it is actually the name of the road. I say this based on your changeset comment.

Enjoy mapping

Hi,

You seem to know Solwezi. Can you check the Solwezi River? I think it has been mapped as 1 river by earlier mappers but is in fact 2 rivers with a confluence to the south of the town and the T5, with both branches flowing south. The Northern Zambian border with the Congo follows the watershed between the Congo basin and the Zambezi basin so rivers here will flow south.

Regards

Hi Frans,

This is actually water left over from the flooding of Zambezi so should be tagged as water with intermittent = yes. It is surrounded by wetland, the green.

Have a look around this school node/6497414967. I have tagged it as a school because it shows all the signs of a rural Zambian Government school. The buildings around a square the small circular outline where the flagpole goes (under the node), but most importantly the dirt athletics track nearby, in this case the light green area to the south east. Now how do you classify the road out of the school? It goes to the athletics track as quite a substantial road and then?. That road would have been cleared by the pupils under instruction from the head teacher hence the size, but all other access is via small tracks made by the children walking to school.

Best regards

I can see where you are coming from but I can assure you maybe 1 vehicle a year will pass along theses roads in fact they are more paths than even tracks. These villagers will survive on fishing, note there are only small agricultural fields here. They will dry the fish on racks pack it in sacks and probably travel by bicycle to the nearest centre. There they will either sell it or travel on an old pickup to the next larger centre and so on. Zoom out and have a look at the whole area you will see oxbows and wetlands, this area floods in the wet season, the villages are on the high ground in fact some of the roads and tracks may not even exist any more or may travel in a slightly different line. I have read the comments re mapping in Africa and I still think you and RAytoun are trying to apply 1st world ideas to 3rd world situations. I work on the principle of what is on the ground and not what people feel it should be. Even in your most wild and remote safari are you likely to visit a place such as this, even the aid organisations don't visit here, not even with their nice new white Toyota Land Cruisers.

The talk of driveways and service roads to the settlements really makes me chuckle. Roads through these settlements will have no real outline, just a large bare area with perhaps a tree under which you can park.

I think you must tag as you feel best. I will still tag the way I feel is most appropriate.

Best regards

Hi Frans,

I have looked at the wiki and there is nowhere where it suggests that a through going highway can not be a track. It does give agriculture as an example but adds "that are suitable for two-track vehicles, such as tractors or jeeps". From the unclassified highway wiki this is mentioned " Unclassified roads are considered usable by motor cars" these highways I have tagged as tracks are not useable by motor cars, perhaps pickups and utility vehicles with a high clearance yes. Further down in the page this is stated "UK: unsurfaced country roads ("green lanes") should be tagged as highway=track if they are unfit for standard motorcar usage but available to 4x4, bikes and foot usage", while this is not the UK the same principle applies. From what I understand about OSM is local mappers have precedent when deciding what applies. I'm afraid I don't have a lot of faith in HOT projects as I have found many errors mapped by these teams mainly because they do not understand the situation on the ground and are applying 1st world thinking or stereotypes from what they think Africa is like. Also Africa is not one country so the same classification does not apply across the whole continent. As examples I have found termite mounds mapped as buildings, long burnt stretches of bush tagged as motorways, seasonal wetlands, known locally as 'dambos' tagged as water. These dambos will be completely dry from August through to November and the grass will burn giving a black appearance in the imagery. Most satellite imagery is taken during the dry season, May - November, as there is no cloud cover that is why the irrigated farm fields stand out as bright green circles, everything around is dry.

Have a look at this link to get some idea about mapping in Zambia osm.wiki/WikiProject_Zambia#Mapping

Best regards