Liquefaction Risk in Kathmandu: Clear Study Summary Part II

July 8, 2026 | By Thethoughtpole

This second article continues the work of Pokhrel and colleagues from the University of Bristol and partner institutions, who studied liquefaction potential in Kathmandu Valley under the EPSRC‑funded SAFER project.
Using the Sonmez liquefaction potential index at 75 SAFER/GEO‑591 boreholes and ordinary kriging, they created valley‑wide maps for different earthquake levels and for wet vs dry seasons.The maps show that northern sandy formations (Gokarna) and river plains have higher liquefaction potential, especially in monsoon with shallow water tables, while the southern clayey Kalimati Formation usually has lower potential.
The limited liquefaction observed in the 2015 Gorkha event matches the modelled dry‑season maps, not a low long‑term hazard.

For Full Research Paper Visit https://link.springer.com/article/10.1007/s10518-021-01198-7

Liquefaction potential for the Kathmandu Valley, Nepal: a sensitivity study

1. From Point Data to Valley Maps

Once the authors had PL values at 75 borehole locations, they needed full maps.
Single points cannot show patterns for the whole valley.

They used ordinary kriging, a statistical interpolation method.
Kriging estimates PL at unsampled locations by looking at how PL varies between known boreholes.

They produced eight main maps by combining:

2 hazard models: AVERAGE vs AB03
2 probabilities: 2% vs 10% in 50 years
2 groundwater conditions: wet vs dry

Each map shows zones from non‑liquefiable to very high liquefaction potential.

2. Wet vs Dry Season Behaviour

The maps show a strong seasonal effect.

In wet‑season (monsoon) maps, with shallow water table (about 1.6 m), liquefaction potential is much higher.
In dry‑season maps, with deeper water table (about 5.1 m), liquefaction potential drops in many areas.

Under the AVERAGE hazard 2% in 50 years case (PGA ~1.2 g), the wet‑season maps show large areas with high or very high PL, especially north and near rivers.
This is a severe “worst‑case” for liquefaction.

Under the 10% in 50 years dry‑season case, many central and southern parts of the valley move into low or no liquefaction categories.
This pattern matches the relatively limited liquefaction seen in the 2015 Gorkha earthquake, which happened in late April, near the end of the dry season.

3. North vs South and Role of Geology

The maps show a clear contrast between north and south.

The Gokarna Formation in the north has sand‑rich soil and higher PL (more likely to liquefy).
The Kalimati Formation in the south has clayey silt and lower PL (less likely to liquefy).

River plains and floodplains are consistently mapped as high liquefaction potential zones.
These areas are low‑lying, near water, and have the right sandy or silty layers.

Some central valley spots between Kathmandu and Lalitpur, close to the river, show moderate instead of high potential, especially in dry conditions.
This likely reflects mixed silty–clayey deposits and less saturation at depth during dry season.

4. How Do These Maps Compare With Older Ones?

Older liquefaction maps, such as those linked to the UNDP/UNCHS 1993 project, used fewer boreholes and simpler rules.
They often marked most near‑river areas as high liquefaction potential without a detailed look at soil type and newer data.

The new maps:

Use the SAFER/GEO‑591 database with many more investigations.
Include season (wet vs dry) and several hazard levels.
Show a more detailed pattern: highest risk in northern sands and floodplains, lower risk in southern clays.

This better matches the known geology and CPT/SPT results from more recent studies compiled by the same research group.

5. Correcting Common Misunderstandings

Misunderstanding 1: “2015 proved liquefaction is not a problem here.”

The authors show that 2015 fits a dry‑season, lower‑PL scenario, not a low‑hazard valley.
If a similar or larger quake hits during monsoon, liquefaction could be much more widespread.

Misunderstanding 2: “All valley soils behave the same in an earthquake.”

The maps and borehole data show big differences:

Sandy formations and floodplains → higher liquefaction potential.
Clayey formations and terraces → lower liquefaction potential (though some loose silts still need study).

Misunderstanding 3: “Old maps are enough for planning.”

Older maps do not fully use newer data or capture season and geology effects in detail.
The new PL‑based maps from Pokhrel and co‑authors give a more realistic picture for modern planning and risk models.

6. What This Means for Planning and Safety

The authors see their maps as regional tools, not final answers for each plot of land.
Soils can change quickly over short distances, so important projects still need site‑specific studies.

They recommend:

Including season‑dependent liquefaction in loss and risk assessments.
Collecting more borehole and lab data, especially in poorly studied areas like parts of Bhaktapur and the southernmost valley.
Doing more detailed site response analyses to improve estimates of cyclic stress ratio.
Studying liquefaction of silts in more detail, since many central urban areas rest on silty deposits.