TerraMaris: Plans, Progress And Setbacks Of Atmospheric Research In Indonesia

By: Emma Howard 

To some of us weather enthusiasts, there’s nothing more exciting than a good tropical thunderstorm. For the best storms, you need a good source of humid air from a warm ocean and a hot land surface. If you can find some mountains to push air upwards and initiate convection (the intense vertical motion of air in updrafts and downdrafts which drive storms) all the better.

Figure 1: Development of convection offshore of West Papua. Photo credit: Megan Howard 

As a volcanic archipelago centred right on the equator, Indonesia has all of this and more. So it’s no surprise that Indonesia is the largest of the three major tropical convective hotspots on Earth. Local lore says that rain comes like clockwork during the wet season, occurring every day at the same time for weeks on end. This is borne out in quantitative rainfall observations, which show that after forming over mountains and land during the mid-afternoon and evening, storms tend to move offshore, with regular night-time and early-morning showers over the oceans and seas adjacent to islands. At present, most atmospheric forecast models (which parameterise atmospheric convection rather than resolving it) don’t represent these diurnally propagating systems very well. This makes it challenging to use these models to predict the timing and intensity of convection in Indonesia.

Unfortunately, some of the more intense thunderstorms can have severe impacts on local communities, particularly when associated with large-scale forcing such as Tropical Cyclone Seroja, which struck Timor-Leste and the Indonesian Nusa Tenggara provinces just two weeks ago. Beyond their immediate impact, these storms have subtle impacts further afield. By condensing water vapour into ice and liquid miles above the earth’s surface, intense storms cause latent heat to be released in the upper atmosphere. This heat source drives the Hadley and Walker cells, global scale atmospheric circulation systems which influence weather and climate across the world, including the UK. For these reasons, scientific research into the convection that occurs in thunderstorms in Indonesia is critical for our understanding of the Earth system and improving climate models.

TerraMaris is a large, collaborative research project that is furthering scientific understanding of atmospheric convection in the Indonesian region. The project involves researchers from three UK universities (East Anglia, Reading and Leeds), the UK Met Office and Indonesia’s weather and space agencies (BMKG and LAPAN). TerraMaris aims to transform our understanding of convective processes in Indonesia and their interactions with the largescale flow through an intensive observational and modelling campaign focussed on the circulation systems associated with the daily development and offshore propagation of convection.

Thankfully, the modelling component of our project hasn’t been so affected by the pandemic and is chugging away as normal. We’re generating a set of very high-resolution model simulations over the whole of Indonesia that are able to (at least partially) resolve the convective updrafts and downdrafts in the daily-repeating storms. Unlike many lower resolution models, these simulations are capable of accurately simulating offshore propagating convection. We intend to run 10 simulations, with one during the long-awaited field campaign season, covering the entire December – February rainy season. A wide range of weather conditions will be represented in this sample, and we’ll be able to study the simulated thunderstorms during all of them.

We are able to compare the role these storms play in heating the upper atmosphere to that in more conventional, lower resolution models, which aren’t able to resolve the updrafts and downdrafts and instead have to parameterise them. These models generally don’t represent Indonesian convection very well. It’s early days, but we’re finding that there’s a lot more variability in the height above the ground where heating occurs in the high-resolution models than the low resolution models. Our high-resolution models also simulate the daily formation of storms in the afternoon/evening and their overnight propagation into the oceans really well (see video).

Figure 2: Mean diurnal cycle of precipitation in early TerraMaris simulations.

Because interactions between the atmosphere and the warm tropical oceans are really important in this part of the world, we’re using a carefully designed coupled atmosphere-ocean model to run all these simulations. Full ocean models are very computationally expensive to run, so we’re using a multi-column KPP ocean model in order to simulate turbulent vertical mixing in the near-surface mixed layer. This is the oceanic process that interacts most strongly with the atmosphere, as it transports heat and freshwater fluxes from the atmosphere at the sea surface further down through the upper ocean. The role of ocean currents and other processes are represented by imposing “corrective” sources and sinks of heat and salt which ensure that in the long run, our simulated ocean matches up with observations of the real ocean.

We’re hoping that these simulations will be able to answer some really fundamental questions about how large-scale weather conditions modulate the vertical distribution of convective heating and how important the daily propagating systems are for providing the heat that drives global circulation. This will be useful for improving the representation of Indonesian convection in lower resolution models. If we can improve that, we hope that weather forecasts will improve both locally in Indonesia and globally through interactions with the Hadley and Walker cells. With any luck, by the time we finally step onto that plane, we’ll know a lot more about the storms that we’re trying to observe than we do now!


This entry was posted in Atmospheric circulation, Climate, Convection, Thunder Storms, Tropical convection and tagged . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *