Earth Watching · 地球守望 · Station 01

The world is not short of AI. It is short of insight.

这个世界不缺 AI,缺少的是洞见啊。

Ten things the planet is doing right now that are not on anyone's dashboard.

十件地球正在发生、却几乎没有被任何团队 Watching 的事。

10Watchings 守望
9 / 6Planetary boundaries · crossed 已越界
0Answers offered 不提供答案
Where this began · 起点
如果你提出的问题,都是围绕你知道的事情打转,你是不是在原地打转? A friend, over breakfast. The sentence did not leave.
The harder half · 更严峻的那一半

She said the sharper part quietly, after the coffee arrived: the real risk isn't asking a bad question. It's that the things which are not intuitive, not urgent this quarter, and quietly the most important — are simply not in anyone's watching. No team owns them. No dashboard shows them. So they are not wrong. They are absent.

更严峻的是:那些不那么直观、却长期最重要的事情,根本就没有在你们团队的 Watching 里。没有人负责,没有看板显示,于是它们不是被判断错了,而是根本不在场。

So: an hour of scanning the planet, not for answers, but for the blind spots. Ten questions. Every source linked, so you can go further than this page does.

所以,这一页只做一件事:扫描这颗星球的盲区。十个问题,全部附上可查证的来源。不为给答案,只为把它们放进视野。

Watching观察什么
Challenge难在哪里
Insight洞见
Key data关键数据

This is not an answer sheet. It is a watch list — the point is the attention, not the conclusion. 这不是一份答案,是一张守望清单;重点是注意力,不是结论。

01Ocean · 海洋

The ocean is quietly running out of breath

海洋正在悄悄失去呼吸

Watching 观察

Dissolved oxygen and vertical mixing — not sea-surface temperature. Warmer water holds less oxygen and layers more stiffly, so the ocean's overturning breath slows.

看的是溶解氧与垂直混合,而不是海表温度。水越暖,含氧越少、分层越硬,海洋的"呼吸"就越慢。

Challenge 挑战

Temperature is easy from orbit; oxygen mostly needs ships and floats. So the variable that decides whether fish, fisheries and coastal economies survive is sampled orders of magnitude less often.

温度卫星就能看,氧气得靠船和浮标。于是决定渔业与沿海经济命运的那个变量,采样密度低了几个数量级。

Insight 洞见

We watch the ocean's fever because it is visible from space. We are not watching whether it can breathe.

我们盯着海洋的"发烧",因为卫星看得见;却没有在看它还能不能呼吸。

>90%

of the Earth system's excess heat has gone into the ocean

地球系统多余热量的九成以上被海洋吸收

~2%

global ocean oxygen lost, 1960–2010

1960–2010 全球海洋含氧量下降约 2%

2024record

highest ocean heat content on record

2024 年海洋热含量创历史新高

Sources IPCC SROCC — Ocean & Cryosphere IPCC AR6 WG1 Copernicus Marine Service NOAA

02Nutrients · 氮磷

Everyone counts carbon. Almost nobody counts nitrogen.

所有人都在算碳,几乎没有人在算氮

Watching 观察

Reactive nitrogen and phosphorus flows — the planetary boundary that is transgressed further than climate, and that runs straight through food, water and coastal dead zones.

看的是活性氮与磷的流动——这条行星边界的超越幅度比气候更大,直接连着粮食、水与近海死亡区。

Challenge 挑战

There is no nitrogen equivalent of a tonne of CO₂e: no common unit, no disclosure standard, no market. What cannot be reported does not get managed.

氮没有"吨二氧化碳当量"这样的通用单位,没有披露准则,没有市场。无法被报告的,就无法被管理。

Insight 洞见

Net-zero is a single-dial story told about a nine-dial planet — and we are turning the one dial we know how to read.

"净零"是一个单表盘的故事,而这颗星球有九个表盘;我们只在拧那个自己会读的。

6 / 9

planetary boundaries assessed as transgressed (2023)

九条行星边界中已有六条被越过

~50%

of applied nitrogen never reaches a crop

施用的氮肥约一半从未被作物吸收

of humanity is fed by synthetic nitrogen fixation

全球约一半人口依赖人工固氮而得以进食

Sources Stockholm Resilience Centre — Planetary Boundaries Our World in Data FAO Soils Portal

03AI & Grid · 算力与电网

The constraint on AI is the substation, not the chip

卡住 AI 的不是芯片,是变电站

Watching 观察

Interconnection queues, transformer lead times, and the water used to cool inference — the physical layer under every model roadmap.

看的是电网并网排队、变压器交付周期,以及冷却推理所耗的水——所有模型路线图底下的物理层。

Challenge 挑战

Compute plans are made in months; grid assets are built in years. The gap is not a procurement problem, it is a planning-horizon mismatch that no single team owns.

算力规划以月计,电网资产以年计。这不是采购问题,而是规划时间尺度的错配,且无人负责。

Insight 洞见

The industry asks "how many GPUs?". The planet is asking "how many transformers, on whose grid, cooled with whose water?"

行业在问"要多少张 GPU",地球在问"要多少台变压器、接在谁的电网、用谁的水来冷却"。

~415TWh

global data-centre electricity, 2024 — about 1.5% of world demand

2024 年全球数据中心用电,约占全球用电 1.5%

~945TWh

IEA projection for 2030 — roughly a doubling

IEA 预计 2030 年将翻约一倍

~2.6TW

of generation and storage waiting in US interconnection queues

美国并网排队中的发电与储能容量

Sources IEA — Energy and AI IEA — Data centres & networks Berkeley Lab — Queued Up

04Soil · 土壤

The thinnest layer we all stand on

我们所有人站着的,是最薄的一层

Watching 观察

Topsoil depth, soil organic carbon and soil biodiversity — the living crust that produces most of what we eat and stores more carbon than the atmosphere.

看的是表土厚度、土壤有机碳与土壤生物多样性——它生产我们大部分的食物,储碳量高于大气。

Challenge 挑战

Soil forms on a millennial clock and is lost on an annual one. Degradation is slow, local and unphotogenic, so it never becomes news.

土壤以千年计生成,以年计流失。退化缓慢、分散、不上镜,于是永远成不了新闻。

Insight 洞见

Soil carbon is the largest terrestrial asset on Earth with no owner, no ledger, and no quarterly report.

土壤碳,是地球上最大的一笔陆地资产——没有所有者,没有账本,也没有季报。

~33%

of the world's soils are moderately to highly degraded

全球约三分之一土壤中度至重度退化

~3×

more carbon in soils than in the atmosphere

土壤储碳量约为大气的三倍

~1,000yrs

to form a few centimetres of topsoil

形成几厘米表土需要约千年

Sources FAO — Global Soil Partnership UNCCD — Land degradation Our World in Data

05Groundwater · 地下水

Water we cannot see, weighed from orbit

看不见的水,被卫星称了重量

Watching 观察

Aquifer mass change, measured as tiny variations in Earth's gravity field by the GRACE and GRACE-FO satellites.

看的是含水层的质量变化——由 GRACE / GRACE-FO 卫星通过地球重力场的微小起伏测得。

Challenge 挑战

Groundwater is invisible, unmetered in most of the world, and drawn by millions of independent decisions. It is the ultimate commons with no gauge on the dashboard.

地下水看不见、多数地区不计量、由千万个独立决策抽取。它是最典型的公地,却没有仪表盘。

Insight 洞见

We learned the world's aquifers were emptying by weighing the planet — not by looking down the well.

我们发现世界的含水层正在被抽空,靠的是"称量地球",而不是低头看井。

21 / 37

major global aquifer systems past sustainability tipping points

全球 37 个主要含水层中 21 个已过可持续临界点

~99%

of Earth's liquid freshwater is groundwater

地球液态淡水中约 99% 是地下水

cm/yr

land subsidence in major basins, largely irreversible

主要盆地地面每年沉降数厘米,多不可逆

Sources NASA JPL — GRACE / GRACE-FO UN-Water — Groundwater USGS

06Circulation · 洋流

A current billions depend on, watched by a handful of moorings

数十亿人仰赖的洋流,由少数几个浮标看守

Watching 观察

The Atlantic overturning circulation — the conveyor that sets European winters and the position of tropical rain belts.

看的是大西洋经向翻转环流:它决定欧洲的冬天,也决定热带雨带落在哪里。

Challenge 挑战

Continuous direct observation began only in 2004, from a single line of moorings. A civilisational variable rests on a thin, renewable-every-few-years research grant.

连续直接观测始于 2004 年,只有一条锚系阵列。一个文明级变量,挂在几年一续的科研经费上。

Insight 洞见

The most consequential number in the climate system has about twenty years of record — and a fragile funding line.

气候系统里最要命的那个数字,只有二十年记录,和一条随时可能断的经费线。

2004

start of continuous AMOC observation at 26°N

北纬 26° 连续观测自 2004 年开始

low / high

IPCC framing: low likelihood, high impact this century

IPCC 定性:本世纪低概率、高影响

billions

of people in monsoon regions exposed to a shifted rain belt

季风区数十亿人口暴露于雨带迁移风险

Sources IPCC AR6 WG1 — Ocean circulation National Oceanography Centre NOAA

07Permafrost · 永冻土

The carbon that is in no one's budget

没有被写进任何预算的碳

Watching 观察

Ground temperature and thaw depth across the permafrost region, where frozen soils hold roughly twice the carbon now in the atmosphere.

看的是永冻区的地温与融化深度——那里的冻土封存的碳,约为当前大气含碳量的两倍。

Challenge 挑战

Most corporate and national pathways treat permafrost emissions as external background. It is a liability held off the balance sheet by convention, not by evidence.

多数企业与国家路径把永冻土排放当作外部背景。这是一笔按惯例被移出资产负债表的负债,而非有证据支持。

Insight 洞见

Every carbon budget quietly assumes the ground stays frozen. That assumption has no line item and no owner.

每一份碳预算都默认"地是冻着的"。这个假设没有科目,也没有负责人。

~1,400GtC

carbon stored in permafrost-region soils

永冻区土壤储碳量(约 1,400–1,600 GtC)

3–4×

Arctic warming rate versus the global average

北极升温速率约为全球平均的三到四倍

~2×

permafrost carbon versus carbon now in the atmosphere

永冻土碳约为当前大气碳的两倍

Sources NOAA Arctic Report Card Woodwell Climate Research Center IPCC AR6 WG1

08Aerosols · 气溶胶

When we cleaned the air, the world got hotter

当我们把空气洗干净,世界反而更热了

Watching 观察

Aerosol cooling — including ship tracks over shipping lanes, which brightened marine clouds until sulphur was cut from marine fuel in 2020.

看的是气溶胶的降温作用,包括航线上的"船迹云"——直到 2020 年船燃硫含量被大幅限制。

Challenge 挑战

Aerosol forcing remains the single largest uncertainty in climate projections, and the sign of a policy's climate effect can be opposite to the sign of its health effect.

气溶胶强迫仍是气候预测中最大的不确定项;一项政策对健康有益,对气候的方向却可能相反。

Insight 洞见

Doing the unambiguously right thing for human lungs removed a shade we never knew we were standing in.

我们为人的肺做了一件毫无疑问正确的事,却顺手撤走了一片自己从不知道正站在其中的阴凉。

3.5 → 0.5%S

global marine fuel sulphur cap, from 2020

2020 年起全球船燃硫上限由 3.5% 降至 0.5%

~80%

cut in shipping sulphur emissions

航运硫排放下降约 80%

2024

first calendar year about 1.5 °C above pre-industrial

2024 年成为首个较工业化前高约 1.5°C 的完整年份

Sources IMO — Sulphur 2020 Copernicus Climate Change Service NASA Earth Observatory

09Materials · 材料

The richest ore body is in the drawer

最富的那座矿,在抽屉里

Watching 观察

Electronic waste as a resource flow: the gold, copper, cobalt and rare earths already sitting inside devices people own.

把电子废弃物看作资源流:金、铜、钴、稀土,早已躺在人们手中的设备里。

Challenge 挑战

The world's fastest-growing waste stream is growing about five times faster than documented recycling. Collection, not chemistry, is the bottleneck.

全球增长最快的废弃物流,增速约为正规回收增速的五倍。瓶颈不在冶金技术,而在回收网络。

Insight 洞见

A device company's largest untapped mine is the installed base it has already sold.

对一家硬件公司来说,最大的未开采矿藏,是它已经卖出去的那些设备。

62Mt

e-waste generated worldwide in 2022

2022 年全球产生电子废弃物 6,200 万吨

22.3%

formally collected and recycled

仅 22.3% 被正规收集与回收

82Mt

projected annual e-waste by 2030

预计 2030 年将达 8,200 万吨/年

Sources Global E-waste Monitor 2024 (UNITAR / ITU) ITU

10Orbit · 轨道

Who is watching the watchers

谁在看着那些看着地球的眼睛

Watching 观察

The continuity of Earth observation itself: ageing missions, funding gaps between satellites, and an increasingly crowded low Earth orbit.

看的是"观测能力"本身:老化的卫星、任务之间的空档,以及日益拥挤的近地轨道。

Challenge 挑战

Climate records only have meaning if they are unbroken. A three-year instrument gap can cost decades of comparability — and debris risk is rising faster than replacement cadence.

气候记录的价值在于连续。三年的仪器空档可能毁掉几十年的可比性;而碎片风险的增速,快过卫星更替的节奏。

Insight 洞见

Every number on this page reached you through a handful of ageing instruments in a crowded sky. Watching is itself a fragile infrastructure — and nobody's roadmap owns it.

这一页上的每一个数字,都来自拥挤天空中少数几台正在老去的仪器。"看见"本身就是脆弱的基础设施,而它不在任何一份路线图里。

~36,000

tracked debris objects larger than 10 cm

被编目的 10 厘米以上空间碎片约 3.6 万个

~1.2M

estimated fragments between 1 and 10 cm

1–10 厘米碎片估计约 120 万个

decades

of comparability lost to a single instrument gap

一次仪器断档,可能损失数十年的可比性

Sources ESA Space Debris Office CEOS — Committee on Earth Observation Satellites NASA Earth Observatory

The eleventh watching · 第十一个守望

The ones we did not think to watch

那些我们连"要不要看"都没想过的

Ten is not a finding. Ten is where attention ran out. These are the candidates already knocking — each one is a next station, a deeper dive, another Wow we have not had yet.

"十"不是结论,只是注意力用完的地方。下面这些已经在敲门了——每一个都可以成为下一个 Station,一次更深的下潜,一个我们还没有过的 Wow。

Starting points Living Planet Index UN Environment Programme NOAA — CO₂ trends Our World in Data

Open the next station 开启下一站

每一次守望,都是一个开始。
Every watching is a start. Mind the missing one.