this week, we're going big on the environment, but can we fight or even understand the changes we're inflicting on the planet? well, we're in iceland, pulling c02 out of thin air. talking of thin air, we're high up in the alps monitoring melting snow. and how do you find out what lives in a rainforest? well, here's one high—flying idea. i'm being swabbed! snow. icebergs. glaciers. continental ice sheets.

if it's white and cold, it's part of the earth's cryosphere... ..and it shouldn't come as a surprise to learn that that is getting smaller. working out what's melting and where is important... ..and up here, nearly 3,500m above sea level, we've come to the high—altitude research station overlooking the aletsch glacier in the alps. switzerland is home to 1,400 glaciers, of which aletsch is the biggest, and it's here that we find groups of scientists developing new techniques to watch the ice and snow and monitor their retreat. glaciers are water reservoirs that we have, which is important for groundwater replation. i think this is the important fact, but also if you lose this mass, you have instability in the mountains, so you have more landslides

ongoing, more erosion. it is also a cause of natural hazards, avalanches, if there's a lot of snow and rock falls and similar things, if there's not enough snow and the ground destabilises. a very important one in terms of the energy transition is the hydropower generation. hydropower generation uses the water that runs down from the mountains — and if there is not enough water, they cannot generate enough power. and if there is a lot of water, a lot of snow, they actually need to know this in advance to plan the generation. for all this, we of course need to know where the snow is, because that is the water of the future. there are two new methods being developed here to not just monitor the surface — how much it's melting, how fast the glacier is moving and so on — but also to find new ways of seeing below the surface. these two top antennas are transmitting antennas, and here we have a set

of receiving antennas. so they're two different radars. this gives us a different, let's say, angle and perspective, and with this we can get an extra information that otherwise we would not be able to acquire. esther and marcel�*s team is using radar to penetrate deep into the snowpack. now, different formations of ice crystals reflect the signals back in different ways, revealing the internal structure of the snow, how deep it is and how dense it is. after several weeks of taking measurements on snow cover across the glacier, the team will spend the next few months evaluating that data to discover whether radar really can shape future studies. but from the ground, you can only see so far. to get a wider view, to get a wider view, you need the second method you need the second method

and go even higher and go even higher up than this. up than this. the space measurements the space measurements are needed to cover the area. are needed to cover the area. it is simply impossible it is simply impossible to cover a large area to cover a large area like the entire alps like the entire alps with ground—based measurements. with ground—based measurements. konrad's project is training konrad's project is training an ai system to predict an ai system to predict what will happen to the snow next. what will happen to the snow next. the system has been trained the system has been trained on images from esa's on images from esa's sentinel—2 satellite network. sentinel—2 satellite network. these are a mix of optical these are a mix of optical photographs that we'd see photographs that we'd see with the naked eye, images with the naked eye, images in the infrared and some in the infrared and some created using longer wavelength created using longer wavelength radio waves like marcel radio waves like marcel methods, then they will no and esther�*s and esther�*s ground—based system. ground—based system. what our model effectively does what our model effectively does is it creates a time series. is it creates a time series. the satellites pass over the satellites pass over every couple of days, every couple of days, and it's like a very slow and it's like a very slow motion video of the surface. motion video of the surface. and we acquire this and we acquire this video and process it. in our case, we have a local snow—depth pattern, and we want to predict, with the help of this sequence and also the new observation, what will be the next snow—depth pattern. now, these techniques are still in the experimental stage.

if the teams can improve their video and process it.

that's made it the hottest planet in the solar system — and while earth's greenhouse effects won't be anywhere near as bad, we have to fight its effects. and adrienne murray has been looking at a promising solution back on our world. or at least...i think it is. adrienne: this isn't some faraway planet. these martian—like landscapes are found in iceland. it could almost be a scene from science fiction. seemingly in the middle of nowhere, these towering machines are guzzling up carbon dioxide, a global—warming gas. so could this much—hyped technology help us fight the climate crisis? mammoth is the world's largest direct air capture and storage facility.

what you see here are 12 of our collector containers. when the plant is fully operational, we'll actually have 72 around the plant. that will enable us to capture 36,000 tonnes of c02 every year. it works like a giant vacuum. each of these units is the size of a shipping container and has a dozen powerfulfans sucking in the surrounding air. they pull in an olympic swimming pool�*s worth of air every a0 seconds. and then, inside, a filter separates the c02. in the atmosphere, its concentration is very dilute. capturing even small amounts requires lots of energy, and mammoth gets both power and hot water from the nearby geothermal plant. for us to do direct air capture effectively and efficiently, we want to make sure that we're

using energy that has a low carbon footprint. some would look at this and say, "hang on, where's the industry?" would it not be more efficient to have one of these next to a factory that's actually making the pollution? carbon dioxide tends to just disperse and diffuse in the air. the effectiveness of direct air capture is not dependent on being located close to industrial emitters. ok, i'll let you show me where the c02 goes. that lowest line here, that's actually the c02 that's coming from those 12 containers outside. these two balloons are a really good visual representation of what, in total, one tonne of c02 looks like.

this tower then works a bit like a sodastream, dissolving the pure c02 in water. from the top, we have water coming in, so, like a shower. it's sort of, at home, if you're making sparkling water, same idea. that fizzy water is sent to these igloo—like domes. so here we have one of our injection wells. please come inside. this well is going 700m down into the underground here. the c02 and water is pumped deep into the basalt bedrock, where it reacts and turns to stone. so you've got a couple of rocks there. exactly, yeah, i'm a geologist, so i brought rocks. this is a fresh basalt here, actually, from one of the last volcanic eruptions here in iceland. you can see there's a lot of porosity in here. here you can see there's a lot of these pores now filled with whitish specks. some of these contain the mineralised c02. and carbfix says that happens pretty quickly. we're not talking about millions or tens of thousands of years. around about 95% of the c02

was mineralised here within two years in the pilot project. this is incredibly fast, geologically speaking. mammoth is climeworks�* second commercial plant and almost ten times bigger than the last one, collecting 36,000 tonnes of c02 annually, about the same amount as taking 8,000 petrol cars off the road. but it costs a whopping $1,000 to remove just one tonne. what do we mean by removing emissions? among its customers are microsoft, h&m and lego. worldwide, more plants like these are on the way, though they'll still only remove a fraction of what's needed. and despite calls to slash our emissions, the c02 we churn out continues to grow. do you think direct air capture can be an effective tool for removing carbon? we release about a0 billion tonnes of carbon dioxide to the atmosphere every year, so this won't make a dent into the big problem.

but i think you should use all methods and methodologies to fight this problem. what's it going to take to scale it up, bring costs down and make it really impactful? by the end of the decade, we want to be at a cost of capture of $300—400. technology improvements will help drive down costs. a second lever will be scale. the team says this is just the beginning. mammoth will soon be dwarfed by another, much bigger plant, project cypress in the united states, which will eventually capture a million tonnes of c02 each year. from mammoth to cypress, we're now looking to break that hundreds of thousands of tonnes of capture capacity a year. i really do believe direct air capture and other engineered solutions are going to be able to get us to the point that we need to to help

fight climate change. this apple �*s vision pro has launched six years after it launched six years after it launched in the us, augmented reality, the headset is controlled by eyes, hands and voice. it is also available in australia, canada, france and germany for the first time. displays a really fine, almost like being in real life so one of the things that is really amazing is a cinema experience because even better than the best cinema, really. that is also mean that you really want to watch and there isn't a huge amount of content that has been made specifically for the headset as yet.- headset as yet. china 's biggest _ headset as yet. china 's biggest electric - headset as yet. china 's biggest electric car - headset as yet. china 's l biggest electric car maker headset as yet. china 's - biggest electric car maker b y d is continuing to expand

abroad, reaching a billion—dollar deal to build a manufacturing plant in turkey are. it will be b y d �*s second factory in the seven region with the first being in hungary. have you ever wondered where to recycle your rubbish? google has teamed up with swedish start—up to produce an ai powered recycling app. loaded up on your smart phone, point your camera at the object you want to recycle like a bottle or newspaper and the app will recognise it and then tell you how best to dispose of it. how is al going to affect the uk over the next ten years? while attendees at the future of britain conference, organised by the tony blair institute for global change, don't possess clairvoyant powers, they are very well placed as some of the best informed people in science, the arts and government to try and help answer that question. in a speech opening the event, former prime minister tony

blair made it clear he believes ai is key to the uk's economic success. it will change everything, and positively, if we embrace it in a spirit of innovation, not introspection. and new health secretary wes streeting says that al can play a role in the nhs. doubling the number of ct and mri scanners, and notjust more scanners, but ai enabled scanners, those are massive productivity gains. as founder and ceo of pioneering ai company google deepmind, sir demis hassabis has been developing ai for health care, scientific research and consumer use long before it was fashionable. it will transform everything, so it will be at least as big as the industrial revolution, possibly, you know, bigger, more like the advent of electricity or even fire. what does he think the uk's new labour government should do to make the best of artificial intelligence? i think it's a big opportunity, actually, for the country

and the new government to embrace the economic opportunities that i think are going to come with al and also the scientific possibilities. so, helping with health care, drug design, but also climate. now, the new labour government has said that it wants to be a green superpower, and we know that, at the moment, this technology, artificial intelligence, is very, very power hungry. yeah. so how is that relationship going to work out in the future, do you think? i think, in the short term, that's correct, there's a lot of power needed to scale these systems up. but most of the large companies, including google, have committed to green power usage and net—zero targets. the types of systems we're building, we're training on these large data centres will have enormous implications and good use cases for climate. so, things like getting more efficiency out of existing infrastructure like power grids. also, we're applying ai itself to save power

in the data centres. so, i think, in the end, that what these ai systems do, you know, will save energy and power far outweighing what they're currently using. people worry about their jobs and job security. how will ai affect employment? ithink, like previous big, uh... ..sort of technological revolutions, like the internet or mobile, i think it's going to also create many newjobs, new types ofjobs that we can barely imagine today. there's a lot of american investment, there's a lot of investment going on in china. where does the uk sit? yeah, i think if you look at it, of course, there's the us and there's china, the kind of two superpowers — they have the most investment, the most research. but i think, if you look at the next tier, the uk is right in there in the mix. so, partly to get the economic gains and the prosperity from that, but also to influence how it goes on the world stage. 0k, crystal—ball moment —

how far away do you think we are from artificial general intelligence? so, agi is a term that we sometimes use in the field to describe roughly human—level ai systems that are capable of general solutions to problems, a bit like the human mind can do. and i think we're, you know, maybe... i wouldn't be surprised if it was sort of in the next decade, so perhaps a 50% chance in the next decade. sir demis hassabis, thank you very much indeed. thank you. life is brilliant! it sprung up in almost every part of the planet, in every shape imaginable — and many that aren't. there's a lot of life out there still left to be discovered and described. and that's a fact — most of the world's flora and fauna are as yet undocumented. but that also means that we don't know what effects we humans and our ever—expanding civilisation are having on most of the world's species.

if we start to lose this diversity, you know, and things can't be moved around and they can't adapt, then we sort of have a house of cards that's falling apart. and if we lose the key species in that, then we know we won't be able to restore our ecosystems. cue xprize rainforest — a five—year competition challenging teams from around the world to develop tech to find out what lives where. first prize? $5 million. the mission — to design robots that can autonomously collect environmental dna, edna, from an area of rainforest and also to develop new techniques to analyse the biodiversity contained within that data. environmental dna, simply dna that's being shed off of every species. and our technologies are advanced enough now that we can just take

a water sample or an air sample or a surface sample, and from that we can describe what species are living in a place. biodivx from eth university in zurich is one of the teams that made it through the semifinals in singapore and have been practising for the finals in the amazon rainforest in brazil. and today i've come to zurich to meet their autonomous drone. actually, what's important here is what's hanging underneath the drone. that is a probe on about 50m of wire that can be lowered all the way through the vegetation down to the ground, swabbing dna samples as it goes. the swab is a simple piece of lint—free cloth that drags against the vegetation. and in case you're

wondering why staying above the rainforest and swabbing down through it is the best idea, well, you've obviously never been to a rainforest before. they are pretty tough conditions, so rainforests have very extreme weather conditions, so it can be very hot, it can be very humid, so that's quite of a challenge for the sensor on board of the drone, for the camera. it can be challenging to detect obstacles, and then what is very problematic is to fly eventually inside the rainforest. so this is where it's, you know, extremely cluttered environment with many obstacles and so on. and so the way we solve this problem is to stay above the canopy, butjust having the probe that goes inside the canopy. so in this way we are safely up, but we can still collect data from inside the vegetation where things are interesting and important to be monitored. the drone uses 3d scans of the canopy to work out how close is too close to the tree tops, and if the probe gets stuck, the drone is programmed to first wiggle it up and down, then to give it a sharp tug, and if all else fails,

to cut the cord and return to base. but what's special about this probe is that it allows us to access the canopy, which is very hard. that's 20, 30, even 40m off the ground, and that's a place where we don't know very much about biodiversity. and so the ability to go in and non—invasively sort of swab this space and collect that dna is amazing, because now our other methods are to put up a poisonous gas and fog the tree and all the insects fall out dead. this is, you know, maybe not what we want to do when we just want to know something's there. yeah, you're decreasing the biodiversity right there and then, aren't you, right? yeah. i'm being swabbed! a lot of chemicals in this hair! that's going to skew the sample! each team has 2a hours to collect as much edna as possible from one square kilometre of rainforest and then 48 hours to analyse the samples — and given that most of the millions of these samples collected will still be unidentifiable, teams are also scored on how well they can communicate the diversity

of life that they find. and for biodivx, that includes turning dna into music. and so i got fascinated by this process of being able to communicate what it is that we're finding, because, as i said, most of the things, we don't even know what species it is, we can't even put it in our tree of life. and so how do you communicate to people about things that they can't see and they don't know are there and we have no name for it? well, if you give it a soundtrack, then it's a wonderful way to allow people to connect with biodiversity when we don't have, you know, other means of communicating about what it is. is it a case, then, that in ten years' time,

you might revisit the same forest and the soundtrack might have changed... yeah. ..and there's a danger it might have got more reduced in its variety? mm—hm, yeah. i have this sort of dream that, you know, if species are there in a place, and they've all been evolving and interacting and they're in harmony, that the music from all of them together should sound good, like a symphony, right? and if it's a system that's... you know, maybe the habitat�*s been degraded or an invasive species has come in, you would be able to hear that distortion because it would no longer be in harmony. the dna sequences determine the order of the notes, but then a human composer arranges the track into something that sounds good and brings the variety of life to life. and in fact, this music you're hearing right now? well, it's composed from the dna of a honeybee, a bioluminescent snail and the silver timon tree. calm electronic music plays the xprize rainforest finals happen this summer in the amazon.

and that's it for our dive into nature. hope you've enjoyed it, thanks for watching, and we'll see you next time. hello. fireworks displays on saturday evening at least added a little bit of colour into what was otherwise a pretty grey day. if we look at the satellite imagery from saturday afternoon, there's the uk. lots of cloud in place so, with winds coming in from a south or southeasterly direction, look what's coming, still to come our way. where we do see a few breaks north sea but more particularly to the northeast of scotland, where temperatures drop the furthest to take us into sunday morning. it's around the northeast

where we could see temperatures aberdeen four, lower than that in the countryside. a touch of frost around. most mild enough, though — 8—12 degrees. little change as far as the weather setup is concerned for sunday. high pressure still with us. not a huge amount of wind, so the cloud sits in place. where we have a bit more breeze northern scotland, an enhanced chance of a few more sunny breaks extending into the highlands. still, the northeast of the mainland best favoured. one or two breaks maybe in the west but the cloud thick enough for some patchy rain or drizzle. and later in the day, we could see some sunnier spells towards east anglia and the southeast. temperatures on sunday very similar to saturday. into sunday night we go and again, where we've got the clearer breaks in northern scotland, temperatures will drop the furthest. a mild night. patchy light rain or drizzle. misty over the hills to take us into monday morning. and once again, for most, temperatures not dropping a huge amount to start the day. so, as we go into the start of the new week, not a massive amount will change. a lot of dry weather around. high pressure still dominant. there should be a few more sunny spells as the breeze picks up every now and again but morning fog patches could be an issue. and the other thing — as our area of high pressure

just drifts a little bit further eastwards, we get more of a southerly flow and that, with low pressure out in the west, will help to draw in even warmer air, so temperatures will be back well above normal, particularly across parts of scotland and northern ireland. for monday, though, we start off with that grey scene. a few brighter breaks here and there in the west but there will be a greater chance down towards the english channel coast, the southeast, of seeing some clearer skies move in off the near continent, so an enhanced chance of some sunshine but temperatures still around 12—14 degrees at this stage. with some clearer skies, then, in the southern uk to take us into tuesday morning and light winds, fog patches could be an issue for the commute and where they do form, they could linger for much of the morning, even into early afternoon. elsewhere, lots of cloud to begin with. a few breaks appearing here and there with a bit more breeze in northern scotland. here's probably where we'll see the best of the chances. temperatures again around 12—14 degrees. take care.

live from london, this is bbc news. kamala harris and donald trump criss—cross southern us states on one of the final days of campaigning ahead of tuesday's election. the vice president flew to new york and made a surprise appearance on saturday night live, coming face—to—face with her comedy doppleganger. in other news: lebanon's prime minister calls for an investigation into the kidnapping of a lebanese citizen by israeli commandos. as the clean up continues in spain's valencia, electricity has now been restored in most areas hit by the catastrophic flash floods. and gaming fans from the around the world gather at the first league of legends grand

final held in the uk. hello, i'm samantha simmonds. welcome to the programme. us presidential candidates donald trump and kamala harris are making last—minute pitches to voters ahead of tuesday's election. both have visited the swing state of north carolina with their planes parked alongside each other at charlotte airport. mr trump, the republican nominee, focused on economic issues. ms harris urged people to head to the polls. there are seven states which will decide who becomes the 47th president of the united states. polls show there's very little separating the two candidates — and polls are within the margin of error, so these states could go either way. more than 70 million americans have already cast their vote ahead of tuesday's election, and both candidates have multiple visits planned

to the battleground states over the next two days. let's take a listen to what both