You have deleted a new bypass road that does not exist in the imagery. READ THE NOTES ASSOCIATED WITH THE WAYS BEFORE DELETING. If you are unsure leave the map as it is. HOT projects make a total mess of the map.

Hi OttoR,

I don't want to start a fight but I find tagging of reservoirs with a multipolygon works well and is less confusing than having overlapping ways that have to be separated out when you try to edit the tags and intend to continue in this way. If the road lay out changes and no longer follows the crest of the dam you simply remove the road tag from the dam draw the new way for the road and tag it as such, editing done.

You are making assumptions that there are pipes through the walls of these dams, in fact you are making many many assumptions when mapping in this area. The dam walls are earthen with a compacted clay core to prevent water seepage through them but no earthen dam is 100% water tight and there will always be some seepage this gives rise to the appearance of water on the downstream side of the wall. For overflow of the dam at full supply level there may be a weir, as is the case with the Mwomboshi Dam, or more often a lower section or spillway is to the one end or sometimes both ends of the wall to allow the water to flow at full supply so technically the river course will be over the spillway. The roads on these dams generally run on top of the dam very rarely behind the dam, this is because the seepage makes it boggy, the road will even run down into the spillway and cross over there. Obviously where there is a weir as a spillway the road will not travel along the crest of the wall unless a bridge is built over the weir, but as these are mainly farm dams nobody will go to the expense of constructing a bridge.

As concerns tagging the reservoirs as seasonal, this just adds a complexity that is not necessary, rather draw the way at the high water mark and leave it at that.

Here is a link regarding the completion of the Mwomboshi Dam, https://www.lusakatimes.com/2018/11/10/mwomboshi-dam-in-chisamba-district-completed-9-months-ahead-of-schedule/, you will notice it was completed in 2018 which confirms my suspicion that the satellite imagery is around 3 to 4 years out of date. You will also notice that while the HOT project references Chibombo District this area is in fact Chisamba District and has been so since 2013, admittedly it used to be part of Chibombo District but how were remote mappers to know that.

Another assumption you are making is the number of fences and gates, particularly gates, some may be cattle grids. Also mapping the cleared areas running along side the fences as unclassified roads, they are tracks running inside a cleared firebreak. The seasonal nature of the rains here result in it being very dry from June to November and there being a great fire risk during these times. Many of the dark black areas in the imagery are burnt areas which makes it hard to determine detail when mapping and may lead mappers to tag areas as scrub when in fact they will be a wooded area just burnt off but will be full of leaves in November, the natural vegetation here is what is termed fire climax vegetation ie it is determined by it's resistance to periodic fires.

Regarding the classifying of roads, many many of them are just tracks and even paths that have been formed by mainly foot traffic and small ox carts and bicycles not necessarily vehicles, how many vehicles have you seen in the imagery in these rural areas? They may be used by the odd small pickup truck once in a while. They grow in width as the traffic moves to one side or the other of a wet patch or other obstruction including ant hills that may grow up in the centre of the track and so the track changes slightly over time. Wet patches will become deeper over time so the traffic will start to avoid them, or a thorn tree grows up and the the track moves to avoid it in other words their course is fluid and not a planned road. If you map these as unclassified roads the map will become a dense network of roads and basically unusable.

Just a few of my thoughts on the matter.

I am sure many of these are not actual pitches but there are too many errors to correct individually

C'est la vie in Chisekesi as we say in these parts.

Hi rab,

I make corrections as I know the area fairly well. I make maps for my Garmin device from OSM data and when I spot some glaring errors I come to OSM to try correct them. I have a pet gripe regarding the HOT projects in general as the mappers usually are novices, I know you have to start somewhere but it results in masses of really poor data, not to mention squared off district boundaries and dragged nodes something I am sure would not happen in Europe or other developed parts of the world. The real problem is the validators are not necessarily much more knowledgeable than the mappers. Also after an enthusiastic start excitement dies down and the mess is left. I know they think they are helping but like many of the aid projects in this part of the world the people providing the aid feel good but the benefits on the ground are negligible at best and harmful in many situations. Chibombo District is a favourite for the white Land Cruiser aid organizations as it is close enough to Lusaka that they can be home with their satellite TV and pubs and restaurants after a hard day in the rural areas. You don't find as many in the Gwembe Valley or far North Western Province. Although the current project is in connection with Chibombo District much of the mapping is taking place in Chisamba District which was created from part of the Chibombo District quite a few years back., much of what is being mapped is historical as the satellite imagery is many years out of date, much of it in this area at least 2 years out of date and possibly more.

I see you have tagged outline of the reservoir with a fix me. Note satellite imagery in this part of the world can be many years out of date. This reservoir was completed 2 years ago and with the rains this year it is probably close to full but I have not visited it recently. Problem with HOT projects is they are mapping the past and not current situations.

You also state that multipolygons are not necessary for dams or reservoirs, they may not be necessary but they work well as you don't have ways over lapping each other. Furthermore a reservoir is made up of the water body and the dam wall that holds back the water, ideal situation for a multipolygon don't you think? I notice you have mapped some reservoir outlines very precisely, the levels of these reservoirs will fluctuate greatly throughout the year and from year to year, most of them in this area are for irrigation projects and since it only rains from November to March they will be empty by October of the next year and may not fill fully in the coming years. Have a look at the same reservoirs with different imagery and you will see what I mean.

Hi Uliwanne,
these are not tracks or roads but are water ways kept open by the local people who travel across the Lukanga Swamp in canoes. Lukanga is not a true swamp but a flooded grassland as are most of the large wetlands in this part of the world

Hi,

This is part of the Muchinga Escarpment, while it may not be a classic 'cliff', view it in JOSM with the OpenTopMap imagery layer to get a better idea as to how dramatic the drop off is, It is an extension of the East African Rift Valley.

Hi Becky Candy,

This is not water but a seasonal wetland known in Zambia and Zimbabwe as a dambo https://en.wikipedia.org/wiki/Dambo, as you can see from the Wikipedia page they take many forms and are an important natural feature in Zambia. Very hard to classify with the existing OSM tags as during the dry season they are dry grasslands and burn with wildfires, showing up as black areas in satellite imagery, during the wet season, particularly very wet years, may be more than ankle deep in water and have a small stream running through them or to the edge of them. They may also have a shallow pool of water in the lowest point known as a pan, as these pans start to dry out the local villagers will plant vegetable gardens in them and around the edges of them, these are sometimes visible in the imagery. Dambos are very often found at the headwaters of many small streams and these streams only start to run once the dambos have become water logged so often only start running towards the end of the wet season such as now.

Hi,
Welcome to OpenStreetMap. The highways you have mapped are not actually highways nor embankments but ephemeral streams in seasonal wetlands. In this part of the world the seasonal wetlands are known as 'dambos' and during the dry season, May to November will be a grassland that will burn fiercely and so in some satellite imagery will appear dark black. During the wet season especially December to February they will be waterlogged and depending on how wet the season is will start flowing with small streams. Because they are open areas they form some of the easiest routes through the bush and so tracks may follow the edges and even the dry stream courses but only passable during the dry season. Away from the main roads most highways are actually tracks or very well worn paths that are not always passable by vehicle. Be aware most satellite imagery is taken during the dry season when the sky is cloudless, it can also be many years out of date

Hi uliwanne,

Be careful deleting buildings that do not appear in the imagery you are using, such as Bing, it does not mean they are not there as Bing is often out of date. Much of the building details added by Akros Global Health are from ground surveys and the information is used for anti malaria work. If you remove the information these houses may not be sprayed. Better not to remove details you not sure of. Be fully aware that satellite imagery in this region is not always up to date and try to use multiple sources

@all

Generally the thalweg of a river is considered to be the borderline where the river demarcates the border. This becomes a problem where the river is braided or has many small islands or in the case of Kariba where the dam was built after and the course of the river is no longer discernable. When the border between Zambia (Northern Rhodesia) and Zimbabwe (Southern Rhodesia) was decided this was not a serious problem as they were both British colonial possessions, in fact were administered by the British South Africa Company (BSAC) and not by the Colonial Office of the UK until, in the case of Southern Rhodesia, a form of self government was granted in the early 1920s and for Northern Rhodesia administration was taken over by the Colonial Office at around the same time. The 2 territories had differing colonial statuses NR was a protectorate and SR was a colony. After the formation of the Federation of Rhodesia and Nyasaland (Malawi) the border became less important as a political entity and mainly used as a economic border (taxes, subsidies etc) and obviously movement across them was not restricted. After independence of these territories it has become more of a political issue as well as an economic one, particularly on Kariba where commercial fishing for kapenta has meant fishing rigs have been impounded by both sides for fishing in each others territories. The irony is that kapenta was introduced to Kariba during the Federation to boost the economies and nutritional welfare of the local people. The sand islands in the Zambezi have also been a problem as the smaller ones come and go with the floods each season and there have been tit for tat arrests of mainly sport fishermen on the lower stretches of the Zambezi for supposedly crossing the border illegally. After much negotiation it was decided that so long as the anglers stayed in their boats and did not set foot on the islands it would be OK, apart from those larger islands that were officially recognised as being part of a particular countries territory, on these you are able to get out of your boat if you are a national of that country, not many people are prepared to test this though.

@Greg_Rose, @cRaIgalLAn, @rab, @Ruben Morgado

I think you have a problem finding an authoritive source, many of these borders were roughly drawn up in colonial days with an idea to have boundary commissions at a later date, generally the 2 World Wars intervened and then decolonization, independence and civil wars intervened. As the colonial boundaries were arbitrarily decided they are not really recognised by the people on the ground, hence you can see no clear break in things such as fields etc. The Zambia/Malawi border is not totally settled and there was a recent border dispute on the Congo/Zambia border that was arbitrated by Zimbabwe. Many borders, particularly Zambia/Malawi and parts if the Zambia/Congo borders follow watersheds and these need to be surveyed to demarcate the border. Other border points depend on old explorer journals that may have miss identified the names or the points this is particularly the case on the Zambia/Congo border around Lake Tanganyika. So all in all good luck.

Also be aware of imagery offset and seasonal nature of rivers and lakes in this part of the world. Lake currently extremely low due to a number of years of poor rains, imagery may be out of date and differences between imagery providers will mean you can always make changes but they may not reflect the actual situation on the ground

By doing it this way you break the Kariba relation. Can't see why the wall has to be an area. Am reverting it to fix the Kariba relation

Hi uliwanne,

this stream, way/814141015#map=15/-14.2283/29.2725 is flowing in the wrong direction. and this one,way/809503182#map=17/-14.41701/29.29458, is going over a ridge.

Sometime back I changed all the streams in this area to intermittent and seasonal as they don't always flow but had not changed the directions but have since come back to have a more detailed look and have found many that are flowing in the wrong direction or a link between 2 systems has been assumed and the link is across a ridge. If you use more than one image source such as Bing and Maxar as well as OpenTopoMap for contours you can see these errors. Also the larger roads run along the watersheds so this can give a hint as to direction of flow. At this point along way/429605612, way/429605612#map=16/-14.4311/29.1998 I deleted a stream that joined across the watershed. Sometimes you have to zoom right in and examine the imagery carefully to see the water courses.

Hi,

You are making many many errors with stream directions, even with streams going over ridges and guessing their courses. It takes a great deal of time to correct them and even then I am not sure they are correct.

You do realise this is Leopards Hill, the hill that the 64 000 acre Leopards Hill ranch was named after. The road was so named as it lead to the Leopards Hill ranch. The hill was named Leopards Hill either because it resembled a leopard or alternately, according to somebody who grew up on the ranch, because of the large number of leopards that lived on or around the hill. The ranch was owned by the Susman brothers and the area is still known as Susman by the locals in the area.

Hi,

I see you are attempting to rework the Zambia Mozambique border. I would be interested to know the source of your information. I have found that many of the online sources are not entirely accurate. Added to this a number of the borders are disputed anyway and are not respected by the people on the ground as they were arbitrarily decided in colonial times. Where they follow rivers or watersheds the situation is much clearer. There is currently a border dispute between Zambia and the Congo in the area of Lake Tanganyika as well as Lake Mweru The survey work of the commission deciding the Congo/Northern Rhodesia border in colonial times was interrupted by WWI and never resumed. The important areas of the Copperbelt were surveyed as no party wanted the other to have access to copper deposits they were not entitled to, you will see some mines right against the border and there was the issue of the Congo Pedicle to clarify.The line between Mweru and Tanganyika was based on landmarks from journals of old explorers and missionaries some of which were misnamed or unclear. The border between Malawi and Zambia has not been defined to both parties satisfaction.

I don't know how the border between Zambia and Mozambique was decided but do know that even in the colonial days the exact point between Malawi/Zambia and Mozambique was vague and there was a pub there known as The Portuguese Pub, similar to Crooks Corner on the border between Zimbabwe/South Africa and Mozambique.

The famous quadripoint at Kazungula does not in fact exist, the borders of Namibia and Zimbabwe do not meet, this gap is what allowed the new Kazungula Bridge to be built between Botswana and Zambia, if the point had existed the bridge would have had to cross Zimbabwean territory.

Your proposed line crossing the T6/N9 is to far to the south, Bing imagery clearly shows the border crossing where the current line is, much more developed on the Zambian side than the Mozambique one.

Hi Season3,

I would not worry about being too precise with the outline of the riverbanks of the Chobe/Cuando as they will vary depending on the season, prevailing wind direction as can be seen in the different satellite images from different sources. This will be true for all the large rivers here flowing through floodplains/flooded grasslands. Zambezi, Chobe, Okavango, Kafue, Luapula and Chambeshi. What you see in the images is floating vegetation and reedbeds which by nature move.