Nepal · स्थानीय तह
Every one of Nepal's 753 local governments — 6 metropolitan cities, 11 sub-metropolitan cities, 276 municipalities and 460 rural municipalities across 77 districts and 7 provinces — with its own temperature and rainfall normals, its place among the other 752, and a CMIP6 trajectory to 2050. The unit list is the government's own boundary file; the normals are sampled from CHELSA's 1 km climate surfaces at each unit's centroid and ship with the page, so the whole country is here before you click anything.
Explorer
Search by name in English or Nepali, or narrow by province, district, unit type and elevation band.
{{ selNarrative }}
Mean temperature and monthly rainfall, CHELSA V2.1 (1981–2010). Monsoon months shaded.
Every local government in the same district, by annual mean temperature. This one is filled.
Annual mean temperature against this unit's own 1991–2020 average, ERA5-Land reanalysis. Red is above that average, blue below; the trend is fitted from 1979.
Seven downscaled CMIP6 HighResMIP models, fetched live for this unit. Line is the ensemble mean, band is the model spread.
All 753 local governments, elevation against annual mean temperature.
Provinces
Each figure is the mean across that province's local governments — one unit, one vote, so a Terai municipality and a Himalayan rural municipality count equally.
Mean annual precipitation across each province's local governments.
How Nepal's 753 local governments distribute across Terai, hill, mountain and high Himal.
Full index
Sort any column. The search and filters above apply here too — click a row to load it in the explorer.
Method & sources
The 753 units. Names, P-codes, types, districts, provinces and areas all come from one file: Nepal's Common Operational Dataset for administrative boundaries, published on HDX with the Survey Department of Nepal under CC BY-IGO. Its admin-3 layer is the 753 local governments — 6 metropolitan cities, 11 sub-metropolitan cities, 276 municipalities and 460 rural municipalities — and the unit type is read from the P-code itself. Nothing here is matched, guessed or reconciled between sources. The 22 protected-area polygons in the same layer (Chitwan, Bardiya, Khaptad, Dhorpatan and the rest) are not local governments and are excluded.
What a centroid means. Each unit's point is its polygon centroid — or, for the six units whose centroid falls outside their own concave boundary, a point guaranteed to lie inside it. A large mountain unit's centroid sits in its terrain centre, not at its headquarters, so these are point estimates for the unit's geographic centre and not settlement-level station data. Elevation is Copernicus DEM GLO-90 at that same point, served by Open-Meteo under CC BY 4.0.
The normals. Monthly mean temperature and rainfall are sampled from CHELSA V2.1 at 30 arc-seconds — about 1 km — for 1981–2010, under CC BY 4.0. CHELSA downscales reanalysis with an orographic precipitation model rather than interpolating between stations, which is why it is the reference choice for mountain terrain. It is also the openly licensed one: WorldClim, the obvious alternative, is free but licensed for academic and other non-commercial use only. Monsoon share is June–September rainfall as a fraction of the annual total.
Where to be careful. Rainfall above roughly 4,000 m is the least certain figure here. Nepal has almost no gauges at that altitude, and CHELSA's orographic model puts more precipitation on high ground than station-interpolated products do — the high-Himal units read wetter here than they would from a product built on interpolation alone. Temperature at those elevations is on firmer ground, and everything below the mountain belt is well constrained.
The observed record. {{ warmingNote }}
Reanalysis is not the station record. Across the 753 units ERA5-Land gives a median +0.09 °C per decade since 1979, and it rises with the land: +0.05 in the Terai, +0.11 in the hills, +0.22 in the mountain belt, +0.25 in the high Himal — elevation explains most of the variation between units (r = 0.76). That gradient is the same one the Department of Hydrology and Meteorology found in Nepal's gauges, where the Terai warmed slowest and the high Himalaya fastest over 1971–2014. The rates are not comparable, and should not be quoted against each other: DHM's headline 0.056 °C per year is maximum temperature, which climbs far faster in Nepal than the daily mean this page shows. For a station-based number, cite DHM directly.
The projection. The trajectory to 2050 is the full seven-model CMIP6 HighResMIP ensemble Open-Meteo carries — CMCC-CM2-VHR4, FGOALS-f3-H, HiRAM-SIT-HR, MRI-AGCM3-2-S, EC-Earth3P-HR, MPI-ESM1-2-XR and NICAM16-8S, downscaled to between 20 and 51 km, CC BY 4.0. Open-Meteo describes the pathway as "as close to RCP8.5 as possible within CMIP6", so read it as a high-emissions case, not a central estimate. The headline figure compares the 2031–2050 mean with the 2001–2020 mean; single-year differences on model output are noise. This is the one thing on the page fetched over the network, and if the source is unreachable the page says so rather than showing invented numbers.
Sources: Nepal COD-AB, Survey Department of Nepal via OCHA/HDX, CC BY-IGO · CHELSA V2.1 (Karger et al. 2017), CC BY 4.0 · Open-Meteo (CMIP6 projections, Copernicus DEM GLO-90), CC BY 4.0. Nepali unit names are community-contributed, not official. {{ cacheNote }}