Reading fog data from INSAT 3DR

At 7.57 am today, the India Meteorological Department’s Twitter handle posted this lovely image of fog over North India on January 21, as captured by the INSAT 3DR satellite. However, it didn’t bother explaining what the colours meant or how the satellite captured this information. So I dug a little.

At the bottom right of the image is a useful clue: “Night Microphysics”. According to this paper, the INSAT 3D satellite has an RGB (red, green, blue) imager whose colours are determined by two factors: solar reflectance and brightness temperature. Solar reflectance is a ratio of the amount of solar energy reflected by a surface and the amount of solar energy incident on it. Brightness temperature has to do with the relationship between the temperature of an object and the corresponding brightness of its surface. It is different from temperature as we usually understand it – by touching a glass of hot tea, say – because brightness temperature also has to do with how the tea glass emits the thermal radiation: at different frequencies in different directions.

INSAT 3D’s ‘day microphysics’ data component studies solar reflectance at three wavelengths: 0.5 µm (visible), 1.6 µm (shortwave infrared) and 10.8 µm (thermal infrared). The strength of the visible signal determines the amount of green colour; the strength of the shortwave infrared signal, the amount of red colour; and the strength of the thermal infrared signal, the amount of blue colour. This way, the INSAT 3D computer determines the colour on each point of the screen to produce an image like this:

CB is short for ‘cumulonimbus’

According to the paper:

The major applications of this colour scheme are an analysis of different cloud types, initial stages of convection, maturing stages of a thunderstorm, identification of snow area and the detection of fires.

The authors also note that the INSAT 3D is useful to image snow: while the solar reflectance of snow and the clouds is similar in the visible part of the spectrum, snow absorbs radiation of 1.6 µm strongly. As a result, when the satellite is imaging snow, the red component of the colour scheme becomes very weak.

The night microphysics is a little more involved. Here, two colours are determined not by a single signal but by the strength of the difference between two signals. The computer determines the amount of red colour according to the difference between two thermal infrared signals: 12.0 µm and 10.0 µm. The amount of green colour varies according to the difference between a thermal infrared and a middle infrared signal: 10.8 µm and 3.9 µm. The amount of blue colour is not a difference, and is determined by the strength of a thermal infrared signal of wavelength 10.8 µm.

And voila:

Perhaps a more explanatory image might help understand the each colour:

The ‘K’ denotes the temperature differences in kelvin. This image shows three kinds of clouds. A mature cumulonimbus cell, possibly part of a tropical storm, hangs over West Bengal and is visible mostly in red (but whose blue component indicates it is also very cold). Somewhere north of Delhi, flecks of green dominate, indicating a preponderance of lower clouds. Even further north, a the sky is dominated by a heavy, high cloud system that encompasses lower clouds as well.

By combining day and night microphysics data, atmospheric scientists can elucidate the presence of moisture droplets of different shapes and temperature differences over time, and in turn track the formation, evolution and depletion of cyclones and other weather events.

For example, taking advantage of the fact that INSAT 3D can produce images based on signals of multiple wavelengths, the authors of the paper have proposed day and night microphysics data that they say would indicate a thunderstorm impending in one to three hours.

Both INSAT 3D and INSAT 3DR use radiometers to make their spectral measurements. A radiometer is a device that measures various useful properties of radiation, typically by taking advantage of radiation’s interaction with matter (e.g. in the form of temperature or electrical activity).

Both satellites also carry atmospheric sounders. Despite the name, these instruments have nothing to do with sound. Instead, they measure temperature and humidity and study water vapour as a function of their heights from the ground.

Scientists combine the radiometer and sounder measurements to understand various atmospheric characteristics.

According to the INSAT 3DR brochure, its radiometer is an upgraded version of the very-high-resolution radiometer (VHRR) that the Kalpana 1 and INSAT 3A satellites used (launched in 2002 and 2003, respectively).

The Space Application Centre’s brief for INSAT 3A states: “For meteorological observation, INSAT-3A carries a three channel Very High Resolution Radiometer (VHRR) with 2 km resolution in the visible band and 8 km resolution in thermal infrared and water vapour bands.” The radiometers onboard 3D and 3DR have
“significant improvements in spatial resolution, number of spectral channels and functionality”.

The Kalpana 1 and INSATs 3A, 3D and 3DR satellites aided India’s weather monitoring and warning services with the best technology available in the country at the time, and with each new satellite being an improved as well as better-equipped version of the previous one. So while Kalpana 1 had a launch mass of 1,060 kg and carried a early VHRR and a data-relay transponder, INSAT 3DR had a launch mass of 2,211 kg – in 2016 – and carried an upgraded VHRR, a sounder, a data-relay transponder and a search-and-rescue transponder.

India deactivated Kalpana 1 in September 2017, after 15 years in orbit. The INSAT 3A, 3D and 3DR satellites are currently active in a geostationary orbit around Earth, at inclinations respectively of 93.5º, 82º and 74º E longitudes.

Featured image: INSAT-3DR satellite in a clean room, with its solar panel deployed, ahead of launch in August 2016. Credit: ISRO.

Pandemic: Science > politics?

By Mukunth and Madhusudhan Raman

Former Union health secretary K. Sujatha Rao had a great piece in The Indian Express on January 14, whose takeaway she summarised in the following line:

Science, evidence and data analytics need to be the bedrock of the roll-out policy, not politics and scoring brownie points for electoral advantages.

However, we can’t help but be reminded of the difference between what should be and what will be. We all (at least those of us who have been on the same side since 2014) know what should be. But as we’ve seen with the National Registry of Citizens (NRC), the Citizenship (Amendment) Act (CAA) 2019 and most recently the farm laws, our present government doesn’t change its mind.

In the last example, the Supreme Court intervened to stay the laws’ implementation but the mediation committee it put together somehow wound up with most members being known to be sympathetic to the government’s position. So what will be, will be – and this is likely to be true vis-à-vis Covaxin as well.

Prime Minister Narendra Modi has already guaranteed as much by determining to foot the cost of 5.5 million doses of Covaxin using the PM CARES fund, which lies beyond public oversight. The Central Drug Standards Control Organisation also played its part by pushing through Covaxin’s approval on terms no one has heard of – and which no one can therefore falsify.

However, this isn’t a pitch for a nihilist position. When Sujatha Rao writes that the government should prize science, evidence and data more than politics and elections, she is right – but we must also ask why. The government has clear incentives to prioritise politics. By thrusting Bharat Biotech – Covaxin’s maker – to the forefront, Modi can claim his ‘Atma Nirbhar’ and ‘Make in India’ schemes have been successful. Also, two important state elections are around the corner: West Bengal and Tamil Nadu.

These are issues that people, but especially ‘Middle Indians’, have an eye on and according to which they vote. The government has also said it is approving Covaxin because it is concerned with the ‘UK variant’. While no reason can be good enough to justify the use of a vaccine candidate in the population sans data from phase 3 clinical trials, the government has effectively set up Covaxin to be failure-proof: if it works, it works; if it doesn’t, it becomes the fault of the variant.

Taken together, Modi’s biggest mistake here is criminal negligence – for pushing Covaxin in the absence of efficacy data (which leads to a cascade of ethical dilemmas) – especially since there are fewer questions over Covaxin’s safety. And negligence is a difficult case to stick to this party or in fact to many people.

Granted, public-spirited science teachers, communicators and journalists can take it upon themselves (ourselves) to persuade readers as to why Covaxin’s approval is really bad – that though everything may turn out okay, it sets a terrible precedent for what this government is allowed to do, how such unchecked power may wreak deadly havoc in future crises, and ultimately that we become a people okay with settling for less, increasingly blind to the banal incrementalism of evil.

In fact, if the mainstream press manages to forget concerns about vaccine apartheid within the country, the dominant narrative as the vaccine roll-out is a few months in is going to be: “India is doing just fine, thank you very much.”

But while the Modi government’s actions may only be negligent – albeit criminally so – in the domains of public healthcare and ‘scientific temper’, they amount to something more egregious if we include the political dimensions of our present moment as well.

None of this means words like those of Sujatha Rao are unnecessary. We need to never forget what should be, and we need to keep protesting for our own sakes. (“Protests sometimes look like failures in the short term, but much of the power of protests is in their long-term effects, on both the protesters themselves and the rest of society.” – Zeynep Tufekci) If we don’t, this government might pretend even less than it currently does that it is following some rules or guidelines from time to time.

However, limiting our exhortations to insist at every turn that “science is more important than politics during a pandemic” risks playing down the importance and influence of political motivations altogether – as well as assuming that the state machinery will automatically give way to scientific ones when lives are at stake.

A politician’s principal responsibility is not to govern but to win elections; good governance is a means to this electoral end. And the way people have voted for many decades attests to the reality of this incentive. While this claim may not be palatable from a theoretical point of view, consider it empirically: the Indian government has seldom responded to national crises to the detriment of potential electoral gains. Examples of such crises include the 1962, 1971 and 1999 conflicts, the nuclear tests and economic liberalisation. During the Emergency, the government itself embodied this crisis.

More recently, numerous ministers and diplomats urged the India and Pakistan governments to find diplomatic solutions after the Pulwama attack and also after the questionable Balakot airstrike, in early 2019. In previous years, they had been preceded by the disagreeable events of Aadhaar implementation, demonetisation and the Goods and Services Tax. But Modi and his fellows won by a bigger margin in 2019 than they had five years earlier.

This happened partly because his success in elections rests on his impression as the Strongman of India, so his resolutions of choice involve flashy displays of strength and machismo.

Against this background: we need to admit political factors into the conversations we – rather, experts like health policymakers, heads of institutions, epidemiologists, healthcare workers, etc. – have from the beginning, instead of ruing the inevitable influence of politics later, so that we may anticipate it and take advantage of it.

For example, consider the conversation surrounding academic publishing. Academics perform most of the work that goes into publishing an academic paper (research, writing and reviewing). Publishing houses add only marginal value to journals – yet publishers charge exorbitant fees to access the results of publicly funded research once it is published. This is unfair, and many academics have said so.

However, the fact that publishing conglomerates are publicly traded companies whose primary responsibility is to generate profits for their shareholders finds little mention in conversations. In this case, the publishers’ profit-seeking motives are fundamental to the problem at hand – but are often disregarded in the first analysis (what should be) and subsequently bemoaned (what will be). For this to happen once is tragic; for it to repeat itself every few months is wasteful.

Similarly, the nationwide lockdown from March to July 2020 served a political purpose: it was a grand gesture, decisive, appealing to ‘Middle Indians’, in addition to supplying the government a pretext to disband protests against the CAA and the NRC. Just before the lockdown, the public conversation had been centred on what the government should be doing. However, most scientists and economists didn’t engage with the political dimension of this decision.

If we had, we may not have been side-tracked into conversations about weekend curfew versus night curfew, or cash transfers versus vouchers, etc. We would perhaps have recognised that our responsibility is not to operate within the parameters set by the government (“How effective was the lockdown?”) but instead recognise that the government’s decisions are politically motivated – so we can ask “Why lock down in the first place?”

The Wire Science is hiring

Location: Bengaluru or New Delhi

The Wire Science is looking for a sub-editor to conceptualise, edit and produce high-quality news articles and features in a digital newsroom.

Requirements

  • Good faculty with the English language
  • Excellent copy-editing skills
  • A strong news sense
  • A strong interest in new scientific findings
  • Know how to read scientific papers
  • Familiarity with concepts related to the scientific method and scientific publishing
  • Familiarity with popular social media platforms and their features
  • Familiarity with the WordPress content management system (CMS)
  • Ability to handle data (obtaining data, sorting and cleaning datasets, using tools like Flourish to visualise)
  • Strong reasoning skills
  • 1-3 years’ work experience
  • Optional: have a background in science or engineering

Responsibilities

  • Edit articles according to The Wire Science‘s requirements, within tight deadlines
  • Make editorial decisions in reasonable time and communicate them constructively
  • Liaise with our reporters and freelancers, and work together to produce stories
  • Work with The Wire Science‘s editor to develop ideas for stories
  • Compose short news stories
  • Work on multimedia rendering of published stories (i.e. convert text stories to audio/video stories)
  • Work with the tech and audience engagement teams to help produce and implement features

Salary will be competitive.

Dalit, Adivasi, OBC and minority candidates are encouraged to apply.

If you’re interested, please write to Vasudevan Mukunth at science@thewire.in. Mention you’re applying for The Wire Science sub-editor position in the subject line of your email. In addition to attaching your resumé or CV, please include a short cover letter in the email’s body describing why you think you should be considered.

If your application is shortlisted, we will contact you for a written test followed by an interview.

Poverty, psychology and pseudoscience

From the abstract of ‘Why Do People Stay Poor? Evidence on Poverty Traps from Rural Bangladesh’, November 24, 2020:

There are two broad views as to why people stay poor. One emphasizes differences in fundamentals, such as ability, talent or motivation. The other, poverty traps view, differences in opportunities stemming from differences in wealth. We exploit a large-scale, randomized asset transfer and panel data on 6000 households over an 11 year period to test between these two views. The data supports the poverty traps view — we identify a threshold level of initial assets above which households accumulate assets, take on better occupations and grow out of poverty. The reverse happens for those below the threshold.

In the resulting worldview this ‘condition’ imposes on people, it’s tempting to see justification for the existence of pseudoscientific enterprises like astrology. Actually, a faith-based binary like ‘requiring faith’ v. ‘not requiring faith’ may be more appropriate here than a science-based binary (‘scientific’ v. ‘unscientific’), if only to emphasise the presence of faith here over the absence of scientific reasoning. So that is, while I can’t ascertain a causal relationship between conditions like the poverty trap and opaque practices like astrology, there’s enough of a correlation here to understand astrology et al as the means by which people rationalise their shared predicament – a predicament that refuses to be allayed by their own efforts.

For example, astrology could provide social, mental and moral incentives for individuals to believe – without having to know – that they were denied any opportunities because ‘their time isn’t right’ and/or that they will continue to luck out, while social realities instead of the alignment of their stars will ensure this is true in some measure. Such faith could also subdue or redirect individuals’ anger or sense of wrongdoing at forces beyond their control, creating ground for social conditions that tolerate oppression more than it ought to be.

Another observation this paper brings to mind is from the work of Sendhil Mullainathan, among others. Researchers from various fields have reported differences in the way poor people make decisions, compared to those who aren’t poor – as if they were less intelligent. However, this perception arises from a sort of cognitive John-Henryism: that is, just as disadvantaged members of society – like Black people in the US – can incur a physical toll imposed by the need to fight for their rights, poor people incur a cognitive toll brought on by the limited availability of resources and the short-lived nature of good fortune.

This doesn’t mean poor people become or are less intelligent, or anything nonsensical like that. Instead, it means poor people’s priorities are different – for example the need for discounts on products, and to maximise absolute savings over percentage savings – in a way that those who aren’t poor may not find optimal for their needs, and that more tasks compete for their attention when they are short on the resources required to execute all of them. As Alice Walton wrote for the Chicago Booth Review in 2018,

In the Wheel of Fortune–style game, the researchers [including Mullainathan] measured how cognitively fatigued the players became. Logic would predict that rich players would be more fatigued, since they were allowed more turns to make more guesses. Instead, the researchers observed that poor players, having received fewer tries to guess at the answers, were more fatigued, having put more effort into each guess.

In an Angry Birds–style game in which people tried to shoot targets, rich players were given more chances to train a virtual slingshot on a target. Poor players, given fewer attempts, spent longer lining up their shots, and many scored more points per shot than rich players. For all the extra shots rich players had, they didn’t do as well, proportionally. “It seems that to understand the psychology of scarcity, we must also appreciate the psychology of abundance. If scarcity can engage us too much, abundance might engage us too little,” the researchers write.

This toll subsequently compromises future choices, and effectively installs another barrier, or trap, in front of people trying to go from being poor in one resource – money, in poverty’s case – to being rich. Walton offers a few examples of policymakers building on these findings to devise better schemes and improve uptake.

In India, where sugarcane farmers are paid annually after the harvest, farmers’ attention scores were the equivalent of 10 IQ points higher than just before the harvest, when farmers were relatively poor, according to data from the 2013 Science study

Offering subsidies or other incentives when people are more receptive to and have the spare capacity to consider them, such as after a harvest or a payday, may make a difference over the long run. One effort, in Tanzania, asked people to sign up for health insurance at cashpoint locations right after payday, and the timing led to a 20 percentage point increase in health-insurance use.

Introducing cognitive aids can help address the limited capacity for attention that may constrain people in poverty. In one study, it helped to show farmers research regarding the most productive ways to plant their crops. When poor, stressed, and in a scarcity mind-set, farmers had a harder time taking in the information. “This result has nothing to do with the intelligence of the farmers,” writes Bryan’s team. “A fact is only obvious if the observer has the spare attentional capacity to notice it.”

I wonder if the converse could also be true: that when homeopaths, phytotherapists, many Ayurveda practitioners and other quack healers offer dubious ways out of difficult healthcare situations, people who are short on attentional space could be likelier to buy into them in order to free up space for other tasks. If so, governments and activists may also need to consider fighting superstition and pseudoscience in healthcare by ensuring more legitimate outcomes – like visiting the local clinic or being able to procure a given drug – require as little cognitive bandwidth as possible.